Human-Computer Interaction · arXiv:2611.17058 v1 · submitted 12 Nov 2026 · 156 references
Neural Representations as New Modes of Interactivity
The paper says
“Human-AI interaction today runs almost entirely through prompts and outputs, keeping AI systems opaque to general users.”
more of the abstract
We introduce Neural Representation Interaction (NRI), a paradigm that treats the internal representations of neural networks as an interaction material people can inspect, steer, and edit. Situating NRI alongside graphical, tangible, and bio-computer interaction, we argue that the concepts and algorithms models encode internally offer representational affordances for richer human-AI interaction. We synthesize actionable mechanistic interpretability methods and, drawing on General Morphological Analysis, construct a design space spanning Interaction, Target, Referent, and Purpose. Mapping existing interfaces onto this space reveals that most serve expert auditors, while few support lay users. We then present speculative applications in education, personal user modeling, and hiring audits, and discuss design principles, limitations, and sociotechnical risks. NRI offers HCI researchers abstractions for designing interactions that work with AI computation rather than only through its inputs and outputs.
from the abstract
Pith says
Machine review · deepseek-v4-flash · 13 Nov 2026 · what was read was not recorded · not checked by a person
It can be wrong. Challenge this review or reply to it.
Holds up, with conditions. A well-structured conceptual paper that names a real gap—HCI lacks a paradigm for treating model internals as interaction material—but the central claim about lay users is undersupported and the framework's viability depends on MI reliability the paper itself flags as unresolved. This line is wrong
4 major objections · 5 minor
How often objections from reviews like this one are right: not yet measured
Not checked: figures, code, what the 156 references say.
The paper says, at Section 4 / Table 2
The review did not quote this passage.
Pith says: Objection 1 of 4 · major
The central mapping claim that “most serve expert auditors, while few support lay users” (abstract, b5) and that “only TalkTuner, Neural Transparency, Multi-Turn Neural Transparency fall under interfaces for lay users” (Table 2 note, b144) is not auditable. The paper does not provide a codebook for how systems were selected, how “lay user” versus “expert auditor” was operationalized, or how the four-facet assignments were made and checked (e.g., by independent raters).
more of objection 1
Since this empirical claim is a headline result, the authors should add an explicit classification procedure and a justification for the system-selection boundary.
The paper says, at Section 3 (GMA facet selection)
The review did not quote this passage.
Pith says: Objection 2 of 4 · major
The claim that the four facets form a complete and cross-consistent design space is an ad hoc assumption. The paper describes the GMA procedure (extraction from surveys, affinity mapping) but does not report whether any candidate facets were discarded, how the named four were chosen over alternatives, or any validity check for the cross-consistency of the cells. Without such transparency, the design space is a useful heuristic but the completeness claim should be softened or made testable.
The paper says, at Section 5 vs. Section 6
The review did not quote this passage.
A classifier read this sentence of the paper as acknowledging the concern (probability 0.66):
“For instance, probes and steering vectors rest on the linear represen- tation hypothesis, yet not all features seem to be linear [37], mo- tivating ongoing work on manifold-based steering [134]; existing steering vectors generalize inconsistently beyond the distribution they were derived from and produce unintended side effects on unrelated behavior [121].”
Pith says: Objection 3 of 4 · major
Section 6 openly acknowledges that probes/steering rest on the linear-representation hypothesis, that SAEs “may not recover ground-truth and human-interpretable features,” and that verbalization “may risk confabulating descriptions” (b218–b220). Yet the speculative applications in Section 5, particularly Latent Mirror and Causal Auditor, assume faithful causal attribution and reliable intervention without a validation protocol.
more of objection 3
The authors should integrate the reliability uncertainty into the design narrative: e.g., propose concrete validation steps (ground-truth feature checks, causal-scrubbing-style tests, user studies measuring downstream trust calibration) or explicitly frame the applications as contingent on future MI reliability.
The paper says, at Section 1 (representational bits)
The review did not quote this passage.
Pith says: Objection 4 of 4 · major
The term “representational bits” is introduced as a core concept of NRI and used rhetorically throughout the paper, but it is never given an operational definition. As written, it conflates activation values, SAE features, and edit targets. A clear formal definition (or an explicit statement that “representational bits” is a rhetoric-level placeholder for MI features) is needed so that claims such as “users can act on representational bits” are meaningful.
5 minor objections
The paper says, at Section 2.1
Pith says: minor objection 1 of 5
The sentence “There is a large body of evidence that features within LLMs are represented as linear directions” should be balanced with the acknowledged nonlinear-features caveat earlier in the paragraph, so the reader is not left with the impression that linearity is the norm.
The paper says, at Section 3 (cross-consistency)
Pith says: minor objection 2 of 5
The sentence “Parameter have soft incompatibility issues with user information which is often measured with the model activations” contains a grammatical error and an unclear subject; revise for clarity.
show 3 more
The paper says, at Table 2
Pith says: minor objection 3 of 5
The Circuit Tracer row lists Target as “Circuits, Components” but the accompanying text in Section 4.2 also refers to tracing circuits and editing activations; please ensure the table cells match the prose exactly.
The paper says, at References
Pith says: minor objection 4 of 5
References [81] and [82] are the same ROME paper listed twice with slightly different formatting; merge them into a single entry to avoid duplicate citations.
The paper says, at Section 3 (Visualizations)
Pith says: minor objection 5 of 5
The visualization paragraph is informative but abruptly placed after the cross-consistency discussion; it might fit better as a short subsection or be explicitly tied to the four-facet table.
What would settle it
Run a pre-registered adversarial-label study in a TalkTuner-style interface: take a real steering-direction UI on a small open LLM and, for half of the concepts, replace the MI-provided labels with random labels sampled from the same vocabulary; ask lay participants to judge which labels are real by testing each control's visible behavioral effects. If label-discrimination accuracy is near chance, NRI interfaces are operating on plausible descriptions rather than on representations users can genuinely interpret and direct; if participants detect the swap reliably, the concern is mitigated.
from the skeptic's reading
How the review went
4 readings by deepseek-v4-flash · 13 Nov 2026
- Reader holds up, with conditions
- Skeptic agrees
- Editor wrote the verdict above
- Referee major revision: 4 major objections, 5 minor
Not run on this paper: the simulated authors' defence.
Pith did not check this review's objections against the paper.
The review in full
The review in brief, machine-written
Stanley,
This is a serious conceptual paper that does something useful: it names and organizes a design space that HCI researchers have been circling but haven't articulated as a paradigm. The four dimensions (Interaction, Target, Referent, Purpose) are sensible, and the mapping of existing tools onto them is a legitimate synthesis. The paper is honest about its limitations—Section 6 openly lists the reliability problems of probes, SAEs, and verbalization, and the architectural assumptions baked into the framework. That candor is a real strength.
What's new is the framing itself: treating internal representations as an interaction material, analogous to how GUI treats pixels and TUI treats physical objects. That's a genuine conceptual contribution, and the paper earns it by situating NRI against prior paradigms and by showing how existing MI interfaces fit (or don't fit) the space.
Now the soft spots. The biggest is that the central mapping claim—that most interfaces serve expert auditors while few support lay users—rests on a curated table (Table 2) with no systematic corpus or codebook. The classification of which interfaces serve lay users is plausible but not auditable. More importantly, the paradigm's viability for lay users depends on MI outputs being both faithful and interpretable without training. The paper concedes both are open problems (b218–b220), yet the design space has no fidelity or validation dimension, and the speculative applications in Section 5 assume reliable causal attribution and intervention. That's not a fatal flaw for a conceptual paper, but it means the paradigm is currently a promissory note rather than a demonstrated interaction mode.
Minor: the GMA methodology is described but not validated—no inter-rater reliability, no systematic extraction protocol. The paper would be stronger with a reproducibility appendix.
Who this is for: HCI researchers working on interpretability interfaces, and MI researchers who want to understand how their tools might be used by non-experts. It deserves a serious referee—the framing is worth engaging with, even if the empirical support is thin. I'd send it to review but flag that the lay-user claim needs either tempering or a validation plan.
Best, [Your name]
The strongest objection
The central claim that users can 'interpret and direct the neural representations of AI systems' (Section 1) rests on two coupled conditions: (i) MI outputs such as probes, SAEs, and verbalizations correspond to the model's actual computation, and (ii) nonexpert users can read those outputs and act on them without expert training. The paper itself states that not all features are linear, that SAEs 'may not recover ground-truth and human-interpretable features,' and that verbalization 'may risk confabulating descriptions' (b218-b220). This is an acknowledged limitation, but it is load-bearing rather than peripheral: if an SAE label or verbalized description does not track the computation it claims to expose, then an NRI interface is not giving users access to neural representations; it is presenting a plausible-sounding explanation artifact, and 'directing representations' reduces to adjusting a visualization. The paper does not integrate this uncertainty into the framework: the Section 3 design space has no fidelity or validation dimension, the Section 4 mapping to Table 2 supplies no codebook for classifying interfaces or user groups, so the claim that only three systems serve lay users (b144, Table 2 note) is not auditable, and the Section 5 speculative applications assume reliable causal attribution and intervention. TalkTuner's positive user-study result (Section 4.1) is a useful existence proof, but it does not establish the general paradigm.
The reader's weakest assumption names the same hinge: internal representations must be accessible and meaningful to users through MI tools, and the paper concedes the underlying reliability questions. My stress-test agrees and sharpens the point: because NRI is defined by users interpreting and directing representations, MI fidelity is constitutive of the paradigm, not an implementation detail. The paper is honest in Section 6, and the conditional verdict already reflects that honesty. What remains unresolved is that the design space, the Table 2 mapping, and the speculative applications are built as if the reliability question had been settled, with no protocol for validating whether a user-facing operation actually targets the model's computation. The concrete adversarial-label test would directly probe whether lay users can distinguish faithful MI output from plausible filler; that is the minimal empirical condition for NRI to be more than a relabeling of expert interpretability tools.
The referee's report
If the conceptual claim holds, NRI offers HCI a genuinely new paradigm that lets users work with AI computation rather than only through prompts and outputs. The paper's strengths are its careful synthesis of a fragmented MI literature, its explicit and candid enumeration of technical reliability limits (Section 6), and its attempt to give interaction designers a reusable vocabulary (Inspect/Steer/Edit; activations/components/circuits/parameters; concepts/computation/user information/behavior; understanding/sensemaking/decision support/auditing/control). The TalkTuner existence proof is a useful anchor. The paper does not ship new empirical data, but as a conceptual and design-space contribution it is potentially valuable for the HCI community.
Pith's reading of the paper (a machine's summary, not the authors')
Human-AI interaction today runs almost entirely through prompts and outputs, leaving the interior of the model a black box. This paper proposes Neural Representation Interaction (NRI): treating the internal representations of neural networks — the activations, features, circuits, and parameters that encode what the model has learned — as an interaction material that users can inspect, steer, and edit. The authors synthesize mechanistic interpretability into a four-dimensional design space (Interaction, Target, Referent, Purpose) and map existing interfaces onto it, finding that most serve expert auditors while few support lay users. If the paradigm holds, AI stops being an inscrutable oracle and becomes a system whose computation can be questioned, audited, and shaped directly; the paper sketches what that would look like in education, personal user modeling, and hiring audits.
What carries the argument. The central organizing device is the four-dimensional design space derived from General Morphological Analysis, whose facets are Interaction (Inspect, Steer, Edit), Target (Activations, Components, Circuits, Parameters), Referent (Concepts, Computation, User information, Behavior), and Purpose (Mechanistic Understanding, Sensemaking, Decision Support, Auditing, Control). It converts the scattered toolbox of mechanistic interpretability into a combinatorial grid in which each full interaction is a cell, letting the paper characterize existing systems and identify empty cells worth designing for. The material the space operates on is the set of 'representational bits' — the linear feature directions, sparse dictionary features, transformer components, and subgraph circuits that mechanistic interpretability extracts from model activations.
Core claim. The paper's central claim is that the semantic structure inside a trained neural network, not just the text it emits, is an interaction surface. It names the elementary units of that structure 'representational bits' and defines NRI as interaction that enables users to interpret and direct them, placed on the same footing as graphical, tangible, and bio-computer interaction. The case rests on three moves: the affordance logic that produced earlier interface paradigms applies to neural computation, since models encode concepts and algorithms that never appear in their outputs and those encodings can be read, pushed, and rewritten; the technical means already exist in mechanistic interpretability (linear probes, steering vectors, verbalization, sparse autoencoders, attribution, circuit discovery) and can be reframed as user-facing capabilities; and a design space built with General Morphological Analysis — Interaction (Inspect, Steer, Edit), Target (Activations, Components, Circuits, Parameters), Referent (Concepts, Computation, User information, Behavior), Purpose (Mechanistic Understanding, Sensemaking, Decision Support, Auditing, Control) — lets designers see which cells are occupied and which are open. Mapped onto this space, the tools built so far cluster around expert auditing, with TalkTuner, Neural Transparency, and Multi-Turn Neural Transparency as the rare lay-user exceptions. The authors state plainly that the design space presumes a transformer architecture and that the underlying methods are still works in progress.
The premise it rests on. The whole paradigm rests on the assumption that what mechanistic interpretability surfaces — approximately linear feature directions, sparse-autoencoder decompositions, verbalized descriptions of activations — faithfully reflects the model's actual computation and can be read and acted on by people who are not interpretability experts; the paper itself concedes that feature linearity, autoencoder fidelity, and verbalization trustworthiness are all unresolved.
If the paper is right (Pith's inferences, not the authors')
Human-AI interaction gains a fourth interaction paradigm alongside graphical, tangible, and bio-computer interfaces, with representational affordances — the coupling of a model's internal states with a user's ability to perceive and act on them — as its defining principle.
The mapping onto the design space shows that nearly all current interpretability interfaces occupy cells aimed at expert auditors and AI-safety teams, while only a handful (TalkTuner, Neural Transparency, Multi-Turn Neural Transparency) serve lay users; the underexplored cells mark where new NRI tools would sit.
Existing actionable mechanistic interpretability — probes, steering vectors, verbalization, sparse autoencoders, attribution and circuit discovery — becomes a design resource, so building NRI interfaces does not require new model science, only re-presentation of existing methods.
The speculative applications become concrete design targets: a teaching tool that decomposes literary style into sparse concepts, a transparent user model (Latent Mirror) that lets people inspect and edit what the AI believes about them, and a causal auditing tool for hiring that tests whether demographic proxies actually drive recommendations.
Two design principles follow directly: progressive disclosure (reveal model internals in layers of complexity) and iterative hypothesis testing (support users in forming and checking causal questions about the model's behavior).
The same mapping that shows most tools serve expert auditors suggests that the binding constraint on NRI is translational, not technical: the real open problem is re-presenting circuit tracers and feature dashboards as progressive, testable interfaces for lay users.
A testable consequence that follows from the paper's logic but is not tested there: interaction that lets users causally manipulate representations of themselves, as in Latent Mirror, should improve trust calibration and reduce over-reliance more than the descriptive explanations XAI dashboards provided — because the user is editing, not just reading.
The authors' own architectural caveat implies the design space is transformer-relative; a natural extension would be an architecture-relative Target dimension, where recurrent and state-space models map to compressed-state components and diffusion models map to dictionary features with no circuit analogue.
If personalized NRI interfaces succeed, mechanistic interpretability tools would gain a second life as consumer privacy surfaces: letting people see and edit what a model believes about them turns the paper's data-provenance concern into a direct user-facing feature rather than a policy footnote.
Check this review
Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.
How often reviews like this one are right
Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.
How often objections from reviews like this one are right: not yet measured
Reviews by people
This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.
The paper
Abstract
Human-AI interaction today runs almost entirely through prompts and outputs, keeping AI systems opaque to general users. We introduce Neural Representation Interaction (NRI), a paradigm that treats the internal representations of neural networks as an interaction material people can inspect, steer, and edit. Situating NRI alongside graphical, tangible, and bio-computer interaction, we argue that the concepts and algorithms models encode internally offer representational affordances for richer human-AI interaction. We synthesize actionable mechanistic interpretability methods and, drawing on General Morphological Analysis, construct a design space spanning Interaction, Target, Referent, and Purpose. Mapping existing interfaces onto this space reveals that most serve expert auditors, while few support lay users. We then present speculative applications in education, personal user modeling, and hiring audits, and discuss design principles, limitations, and sociotechnical risks. NRI offers HCI researchers abstractions for designing interactions that work with AI computation rather than only through its inputs and outputs.
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.
abstract compiled · 2026-11-14
The authors' main claim, as the classifier picked it out 0.76
“NRI offers HCI researchers abstractions for designing interactions that work with AI computation rather than only through its inputs and outputs.”
How the main claim is stated: asserted 0.96
What the abstract itself states, as probabilities: claims first 0.11 · states limitation 0.85 · human participants 0.14 · controlled comparison 0.03 · randomized assignment 0.01 · states quantitative result 0.04 · compares with prior results 0.42 · mentions code or data availability 0.03
The abstract, sentence by sentence
This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.
| Sentence | Role | The authors' own work | States the main conclusion | Strength |
|---|---|---|---|---|
| Human-AI interaction today runs almost entirely through prompts and outputs, keeping AI systems opaque to general users. | background 0.93 | 0.09 | 0.03 | asserted 1.00 |
| We introduce Neural Representation Interaction (NRI), a paradigm that treats the internal representations of neural networks as an interaction material people can inspect, steer, and edit. | contribution 1.00 | 0.89 | 0.20 | asserted 1.00 |
| Situating NRI alongside graphical, tangible, and bio-computer interaction, we argue that the concepts and algorithms models encode internally offer representational affordances for richer human-AI interaction. | interpretation 0.46 | 0.51 | 0.30 | asserted 0.80 |
| We synthesize actionable mechanistic interpretability methods and, drawing on General Morphological Analysis, construct a design space spanning Interaction, Target, Referent, and Purpose. | method 0.66 | 0.93 | 0.07 | asserted 0.99 |
| Mapping existing interfaces onto this space reveals that most serve expert auditors, while few support lay users. | result 0.90 | 0.78 | 0.09 | asserted 1.00 |
| We then present speculative applications in education, personal user modeling, and hiring audits, and discuss design principles, limitations, and sociotechnical risks. | contribution 0.84 | 0.85 | 0.09 | speculative 0.73 |
| NRI offers HCI researchers abstractions for designing interactions that work with AI computation rather than only through its inputs and outputs. | interpretation 0.63 | 0.40 | 0.65 | asserted 0.96 |
What kind of paper the classifier reports this is: position or commentary 0.92 · evidence from argument 0.98
What was examined, and what was not
Sources read
- arXiv listing abstract v1
Not examined
- full text
- figures and tables
- references and citing sentences
- relations to other papers
Levels run
- P abstract
Not run
- S sections
- T body sentences
- C citations
- R cross-paper relations
How the labels were made
Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-14: 1 request, 5,582 input tokens.
Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON
Figures (2)
References (156)
- [1] Gavin Abercrombie, Amanda Cercas Curry, Tanvi Dinkar, Verena Rieser, and Zeerak Talat. 2023. Mirages. On Anthropomorphism in Dialogue Systems. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, Houda Bouamor, Juan Pino, and Kalika Bali (... on Pithdoi:10.18653/v1/2023.emnlp-
- [2] Sylvia Adamson, Lynette Hunter, Lynne Magnusson, Ann Thompson, and Katie Wales (Eds.). 2001.Reading Shakespeare’s Dramatic Language: A Guide. Arden Shakespeare, London on Pith
- [3] Emmanuel Ameisen, Jack Lindsey, Adam Pearce, Wes Gurnee, Nicholas L. Turner, Brian Chen, Craig Citro, David Abrahams, Shan Carter, Basil Hos- mer, Jonathan Marcus, Michael Sklar, Adly Templeton, Trenton Bricken, Callum McDougall, Hoagy Cunningham, Thomas Henighan, Adam Jermyn,... on Pith
- [4] Saleema Amershi, Maya Cakmak, W. Bradley Knox, and Todd Kulesza. 2014. Power to the People: The Role of Humans in Interactive Machine Learning.AI Magazine35, 4 (2014), 105–120. doi:10.1609/aimag.v35i4.2513 on Pithdoi:10.1609/aimag.v35i4.2513
- [5] Eitan Anzenberg, Arunava Samajpati, Sivasankaran Chandrasekar, and Varun Kacholia. 2025. Evaluating the promise and pitfalls of llms in hiring decisions. arXiv preprint arXiv:2507.02087(2025) on PithPith's reviewarXiv
- [6] Georgii Aparin, Tasnima Sadekova, Alexey Rukhovich, Assel Yermekova, Laida Kushnareva, Vadim Popov, Kristian Kuznetsov, and Irina Piontkovskaya. 2026. AudioSAE: Towards Understanding of Audio-Processing Models with Sparse AutoEncoders. InProceedings of the 19th Conference of t... on Pith
- [7] Andy Arditi, Oscar Obeso, Aaquib Syed, Daniel Paleka, Nina Panickssery, Wes Gurnee, and Neel Nanda. 2024. Refusal in language models is mediated by a single direction.Advances in Neural Information Processing Systems37 (2024), 136037–136083 on Pith
- [8] Shuai Bai, Yuxuan Cai, Ruizhe Chen, Keqin Chen, Xionghui Chen, Zesen Cheng, Lianghao Deng, Wei Ding, Chang Gao, Chunjiang Ge, et al. 2025. Qwen3-VL Technical Report.arXiv preprint arXiv:2511.21631(2025). https://arxiv.org/abs/ 2511.21631 on PithPith's reviewarXiv
- [9] Gagan Bansal, Tongshuang Wu, Joyce Zhou, Raymond Fok, Besmira Nushi, Ece Kamar, Marco Tulio Ribeiro, and Daniel S. Weld. 2021. Does the Whole Exceed Its Parts? The Effect of AI Explanations on Complementary Team Performance. InProceedings of the 2021 CHI Conference on Human Fa... on Pithdoi:10.1145/3411764.3445717
- [10] David Bau, Hendrik Strobelt, William Peebles, Jonas Wulff, Bolei Zhou, Jun- Yan Zhu, and Antonio Torralba. 2019. Semantic Photo Manipulation with a Generative Image Prior.ACM Transactions on Graphics (Proc. SIGGRAPH)38, 4, Article 59 (2019). doi:10.1145/3306346.3323023 on Pithdoi:10.1145/3306346.3323023
- [11] David Bau, Jun-Yan Zhu, Hendrik Strobelt, Bolei Zhou, Joshua B. Tenen- baum, William T. Freeman, and Antonio Torralba. 2019. GAN Dissection: Visualizing and Understanding Generative Adversarial Networks. InInter- national Conference on Learning Representations (ICLR). arXiv:18... on PithPith's reviewarXiv
- [12] Leonard Bereska and Efstratios Gavves. 2024. Mechanistic interpretability for AI safety–a review.arXiv preprint arXiv:2404.14082(2024) on PithPith's reviewarXiv
- [13] Leon Bergen, Usha Bhalla, Sidharth Baskaran, Max Loeffler, Raphael Sarfati, Dhruvil Gala, Ryan Panwar, Santiago Aranguri, Thomas Fel, Atticus Geiger, Matthew Kowal, Siddharth Boppana, Daniel Balsam, Owen Lewis, Jack Merullo, Thomas McGrath, and Ekdeep Singh Lubana. 2026. Anato... on PithPith's reviewarXiv
- [14] Umang Bhatt, Javier Antorán, Yunfeng Zhang, Q. Vera Liao, Prasanna Sat- tigeri, Riccardo Fogliato, Gabrielle Melançon, Ranganath Krishnan, Jason Stanley, Omesh Tickoo, Lama Nachman, Rumi Chunara, Madhulika Sriku- mar, Adrian Weller, and Alice Xiang. 2021. Uncertainty as a Form... on Pithdoi:10.1145/3461702.3462571
- [15] Stephen Booth (Ed.). 1977.Shakespeare’s Sonnets. Yale University Press, New Haven on Pith
- [16] Zana Buçinca, Maja Barbara Malaya, and Krzysztof Z. Gajos. 2021. To Trust or to Think: Cognitive Forcing Functions Can Reduce Overreliance on AI in AI-assisted Decision-making.Proc. ACM Hum.-Comput. Interact.5, CSCW1, Article 188 (April 2021), 21 pages. doi:10.1145/3449287 on PithPith's reviewdoi:10.1145/3449287
- [17] Lucius Bushnaq, Dan Braun, and Lee Sharkey. 2025. Stochastic Parameter Decomposition. arXiv:2506.20790 [cs.LG] on PithPith's reviewarXiv
- [18] Aaron Chatterji, Tom Cunningham, David J. Deming, Zoë Hitzig, Christopher Ong, Carl Yan Shan, and Kevin Wadman. 2025.How People Use ChatGPT. Work- ing Paper 34255. National Bureau of Economic Research. doi:10.3386/w34255 Introduces the “Asking, Doing, Expressing” taxonomy; Ask... on Pithdoi:10.3386/w34255
- [19] Haozhe Chen, Carl Vondrick, and Chengzhi Mao. 2024. SelfIE: Self- Interpretation of Large Language Model Embeddings. InProceedings of the 41st International Conference on Machine Learning (ICML). https://arxiv.org/ abs/2403.10949 on PithPith's reviewarXiv
- [20] Jianhui Chen, Yuzhang Luo, and Liangming Pan. 2026. Mechanistic Data Attribution: Tracing the Training Origins of Interpretable LLM Units. arXiv:2601.21996 [cs.CL] on PithPith's reviewarXiv
show 80 more
- [21] Runjin Chen, Andy Arditi, Henry Sleight, Owain Evans, and Jack Lindsey. 2025. Persona vectors: Monitoring and controlling character traits in language models. arXiv preprint arXiv:2507.21509(2025) on PithPith's reviewarXiv
- [22] Valerie Chen, Q. Vera Liao, Jennifer Wortman Vaughan, and Gagan Bansal on Pith
- [23] Yida Chen, Aoyu Wu, Trevor DePodesta, Catherine Yeh, Kenneth Li, Nicholas Castillo Marin, Oam Patel, Jan Riecke, Shivam Raval, Olivia Seow, et al on Pith
- [24] Myra Cheng, Cinoo Lee, Pranav Khadpe, Sunny Yu, Dyllan Han, and Dan Jurafsky. 2026. Sycophantic AI decreases prosocial intentions and promotes dependence.Science391, 6792 (2026), eaec8352 on Pith
- [25] Minh Duc Chu, Patrick Gerard, Kshitij Pawar, Charles Bickham, and Kristina Lerman. 2025. Illusions of Intimacy: How Emotional Dynamics Shape Human-AI Relationships.arXiv preprint arXiv:2505.11649(2025) on PitharXiv
- [26] Wolfgang Clemen. 1987.Shakespeare’s Soliloquies. Methuen, London on Pith
- [27] Benjamin Cohen-Wang, Harshay Shah, Kristian Georgiev, and Aleksander Mądry. 2024. ContextCite: Attributing Model Generation to Context. InAd- vances in Neural Information Processing Systems (NeurIPS), Vol. 37. doi:10.52202/ 079017-3035 on Pith
- [28] Arthur Conmy, Augustine Mavor-Parker, Aengus Lynch, Stefan Heimersheim, and Adrià Garriga-Alonso. 2023. Towards automated circuit discovery for mechanistic interpretability.Advances in Neural Information Processing Systems 36 (2023), 16318–16352 on Pith
- [29] Bartosz Cywiński and Kamil Deja. 2025. SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders. InProceedings of the 42nd International Conference on Machine Learning (ICML) on Pith
- [30] Valdemar Danry, Pat Pataranutaporn, Yaoli Mao, and Pattie Maes. 2023. Don’t Just Tell Me, Ask Me: AI Systems that Intelligently Frame Explanations as Questions Improve Human Logical Discernment Accuracy over Causal AI Explanations. InProceedings of the 2023 CHI Conference on H... on Pithdoi:10.1145/3544548.3580672
- [31] Philip Davis. 2007.Shakespeare Thinking. Continuum, London on Pith
- [32] Yifan Ding, Matthew Facciani, Amrit Poudel, Ellen Joyce, Salvador Aguinaga, Balaji Veeramani, Sanmitra Bhattacharya, and Tim Weninger. 2025. Citations and Trust in LLM Generated Responses. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 39. 23787–23795. d... on Pithdoi:10.1609/aaai.v39i22.34550
- [33] Jacob Dunefsky, Philippe Chlenski, and Neel Nanda. 2024. Transcoders find interpretable llm feature circuits.Advances in Neural Information Processing Systems37 (2024), 24375–24410 on Pith
- [34] Rudresh Dwivedi, Devam Dave, Het Naik, Smiti Singhal, Rana Omer, Pankesh Patel, Bin Qian, Zhenyu Wen, Tejal Shah, Graham Morgan, et al. 2023. Explain- able AI (XAI): Core ideas, techniques, and solutions.ACM computing surveys 55, 9 (2023), 1–33 on Pith
- [35] Upol Ehsan, Q. Vera Liao, Samir Passi, Mark O. Riedl, and Hal Daumé III on Pith
- [36] Nelson Elhage, Tristan Hume, Catherine Olsson, Nicholas Schiefer, Tom Henighan, Shauna Kravec, Zac Hatfield-Dodds, Robert Lasenby, Dawn Drain, Carol Chen, Roger Grosse, Sam McCandlish, Jared Kaplan, Dario Amodei, Mar- tin Wattenberg, and Christopher Olah. 2022. Toy Models of S... on Pith
- [37] Joshua Engels, Eric J. Michaud, Isaac Liao, Wes Gurnee, and Max Tegmark. 2025. Not All Language Model Features Are Linear. InInternational Conference on Learning Representations on Pith
- [38] Seamful XAI: Operationalizing Seamful Design in Explainable AI. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems. doi:10.1145/3613904.3642755 on Pithdoi:10.1145/3613904.3642755
- [39] Alhussein Fawzi, Matej Balog, Aja Huang, Thomas Hubert, Bernardino Romera- Paredes, Mohammadamin Barekatain, Alexander Novikov, Francisco J. R. Ruiz, Julian Schrittwieser, Grzegorz Swirszcz, David Silver, Demis Hassabis, and Pushmeet Kohli. 2022. Discovering faster matrix mult... on Pithdoi:10.1038/s41586-
- [40] George W. Fitzmaurice, Hiroshi Ishii, and William A. S. Buxton. 1995. Bricks: Laying the Foundations for Graspable User Interfaces. InProceedings of the Huang et al. SIGCHI Conference on Human Factors in Computing Systems (CHI ’95). ACM Press/Addison-Wesley, New York, NY, USA,... on Pithdoi:10.1145/223904.223964
- [41] Jerry Alan Fails and Dan R Olsen Jr. 2003. Interactive machine learning. In Proceedings of the 8th international conference on Intelligent user interfaces. 39– 45 on Pith
- [42] Rohit Gandikota, Joanna Materzyńska, Tingrui Zhou, Antonio Torralba, and David Bau. 2024. Concept Sliders: LoRA Adaptors for Precise Control in Diffu- sion Models. InEuropean Conference on Computer Vision (ECCV) on Pith
- [43] Leo Gao, Tom Dupré la Tour, Henk Tillman, Gabriel Goh, Rajan Troll, Alec Radford, Ilya Sutskever, Jan Leike, and Jeffrey Wu. 2024. Scaling and evaluating sparse autoencoders.arXiv preprint arXiv:2406.04093(2024) on PithPith's reviewarXiv
- [44] Kit Fraser-Taliente, Subhash Kantamneni, Euan Ong, Dan Mossing, Christina Lu, Paul C. Bogdan, Emmanuel Ameisen, James Chen, Dzmitry Kishylau, Adam Pearce, Julius Tarng, Alex Wu, Jeff Wu, Yang Zhang, Daniel M. Ziegler, Evan Hubinger, Joshua Batson, Jack Lindsey, Samuel Zimmerma... on Pith
- [45] Sofia Gomes, João Lopes, and Elisabete Nogueira. 2025. Anthropomorphism in artificial intelligence: a game-changer for brand marketing.Future Business Journal11 (01 2025). doi:10.1186/s43093-025-00423-y on Pithdoi:10.1186/s43093-025-00423-y
- [46] Goodfire. 2024. Goodfire Ember: Scaling Interpretability for Frontier Model Alignment. Goodfire blog. https://www.goodfire.ai/blog/announcing-goodfire- ember on Pith
- [47] Google DeepMind. 2025. AlphaFold: Five Years of Impact. Google DeepMind Blog. Accessed 2026-09-06 on Pith
- [48] Asma Ghandeharioun, Avi Caciularu, Adam Pearce, Lucas Dixon, and Mor Geva on Pith
- [49] Patchscopes: A Unifying Framework for Inspecting Hidden Representa- tions of Language Models. InProceedings of the 41st International Conference on Machine Learning (ICML). https://arxiv.org/abs/2401.06102 on PithPith's reviewarXiv
- [50] Kunal Handa, Alex Tamkin, Miles McCain, Saffron Huang, Esin Durmus, Sarah Heck, Jared Mueller, Jerry Hong, Stuart Ritchie, Tim Belonax, Kevin K. Troy, Dario Amodei, Jared Kaplan, Jack Clark, and Deep Ganguli. 2025. Which Economic Tasks are Performed with AI? Evidence from Mill... on PithPith's reviewarXiv
- [51] Sonia Joseph, Praneet Suresh, Ethan Goldfarb, Lorenz Hufe, Yossi Gandelsman, Robert Graham, Danilo Bzdok, Wojciech Samek, and Blake Aaron Richards on Pith
- [52] Sister Miriam Joseph. 1947.Shakespeare’s Use of the Arts of Language. Columbia University Press, New York on Pith
- [53] Wes Gurnee and Max Tegmark. 2023. Language models represent space and time.arXiv preprint arXiv:2310.02207(2023) on PithPith's reviewarXiv
- [54] Bálint Gyevnár and Atoosa Kasirzadeh. 2026. Bridging the Gap in the Respon- sible AI Divides.arXiv preprint arXiv:2603.14495(2026) on PithPith's reviewarXiv
- [55] Adam Karvonen, James Chua, Clément Dumas, Kit Fraser-Taliente, Subhash Kantamneni, Julian Minder, Euan Ong, Arnab Sen Sharma, Daniel Wen, Owain Evans, and Samuel Marks. 2025. Activation Oracles: Training and Evaluating LLMs as General-Purpose Activation Explainers.arXiv prepri... on PitharXiv
- [56] Eoin M. Kenny, Akshay Dharmavaram, Sang Uk Lee, Tung Phan-Minh, Shreyas Rajesh, Yunqing Hu, Laura Major, Momchil S. Tomov, and Julie A. Shah. 2026. Explainable deep learning improves human mental models of self-driving cars. Nature657, 8130 (Sept. 2026), 114–120. doi:10.1038/s... on Pithdoi:10.1038/s41586-026-10950-5
- [57] Dahye Kim and Deepti Ghadiyaram. 2025. Concept Steerers: Leveraging K- Sparse Autoencoders for Test-Time Controllable Generations.arXiv preprint arXiv:2501.19066(2025) on PitharXiv
- [58] Junsol Kim, James Evans, and Aaron Schein. 2025. Linear representations of political perspective emerge in large language models.arXiv preprint arXiv:2503.02080(2025) on PithPith's reviewarXiv
- [59] Sheer Karny, Anthony Baez, and Pat Pataranutaporn. 2026. Multi-Turn Neural Transparency: Surfacing Neural Activations Improves User Calibration to LLM Behavioral Drift.arXiv preprint arXiv:2605.15455(2026) on PithPith's reviewarXiv
- [60] Sheer Karny, Anthony Baez, and Pat Pataranutaporn. 2026. Neural transparency: Mechanistic interpretability interfaces for anticipating model behaviors for personalized ai. InProceedings of the 31st International Conference on Intelligent User Interfaces. 868–884 on Pith
- [61] Siddique Latif, Moazzam Shoukat, Fahad Shamshad, Muhammad Usama, Yi Ren, Heriberto Cuayáhuitl, Wenwu Wang, Xulong Zhang, Roberto Togneri, Erik Cambria, and Björn W. Schuller. 2023. Sparks of Large Audio Models: A Survey and Outlook.arXiv preprint arXiv:2308.12792(2023) on PithPith's reviewarXiv
- [62] Patrick Leask, Bart Bussmann, Michael Pearce, Joseph Bloom, Curt Tigges, Noura Al Moubayed, Lee Sharkey, and Neel Nanda. 2025. Sparse Autoencoders Do Not Find Canonical Units of Analysis. InInternational Conference on Learning Representations on Pith
- [63] John D. Lee and Katrina A. See. 2004. Trust in Automation: Designing for Appropriate Reliance.Human Factors46, 1 (2004), 50–80. doi:10.1518/hfes.46.1. 50_30392 on Pithdoi:10.1518/hfes.46.1
- [64] Haoxuan Li, Zhen Wen, Qiqi Jiang, Chenxiao Li, Yuwei Wu, Yuchen Yang, Yiyao Wang, Xiuqi Huang, Minfeng Zhu, and Wei Chen. 2025. ConceptViz: A Visual Analytics Approach for Exploring Concepts in Large Language Models.IEEE Transactions on Visualization and Computer Graphics(2025) on Pith
- [65] René F. Kizilcec. 2016. How Much Information? Effects of Transparency on Trust in an Algorithmic Interface. InProceedings of the 2016 CHI Conference on Human Factors in Computing Systems (CHI ’16). Association for Computing Machinery, New York, NY, USA, 2390–2395. doi:10.1145/... on Pithdoi:10.1145/2858036.2858402
- [66] Remi Lam, Alvaro Sanchez-Gonzalez, Matthew Willson, Peter Wirnsberger, Meire Fortunato, Ferran Alet, Suman Ravuri, Timo Ewalds, Zach Eaton-Rosen, Weihua Hu, Alexander Merose, Stephan Hoyer, George Holland, Oriol Vinyals, Jacklynn Stott, Alexander Pritzel, Shakir Mohamed, and P... on Pithdoi:10.1126/science.adi2336
- [67] Tom Lieberum, Senthooran Rajamanoharan, Arthur Conmy, Lewis Smith, Nico- las Sonnerat, Vikrant Varma, János Kramár, Anca Dragan, Rohin Shah, and Neel Nanda. 2024. Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2.arXiv preprint arXiv:2408.05147(2024) on PithPith's reviewarXiv
- [68] Johnny Lin. 2023. Neuronpedia: Interactive Reference and Tooling for Analyzing Neural Networks. https://www.neuronpedia.org Software available from neuronpedia.org on Pith
- [69] Jack Lindsey, Wes Gurnee, Emmanuel Ameisen, Brian Chen, Adam Pearce, Nicholas L. Turner, Craig Citro, David Abrahams, Shan Carter, Basil Hosmer, Jonathan Marcus, Michael Sklar, Adly Templeton, Trenton Bricken, Callum McDougall, Hoagy Cunningham, Thomas Henighan, Adam Jermyn, A... on Pith
- [70] Jack Lindsey, Adly Templeton, Jonathan Marcus, Thomas Conerly, Joshua Bat- son, and Christopher Olah. 2024. Sparse Crosscoders for Cross-Layer Fea- tures and Model Diffing. Transformer Circuits Thread. https://transformer- circuits.pub/2024/crosscoders/index.html on Pith
- [71] Millicent Li, Alberto Mario Ceballos Arroyo, Giordano Rogers, Naomi Saphra, and Byron C. Wallace. 2025. Do Natural Language Descriptions of Model Activations Convey Privileged Information?arXiv preprint arXiv:2509.13316 (2025). https://arxiv.org/abs/2509.13316 on PithPith's reviewarXiv
- [72] Yuxuan Liang, Haomin Wen, Yuqi Nie, Yushan Jiang, Ming Jin, Dongjin Song, Shirui Pan, and Qingsong Wen. 2024. Foundation Models for Time Series Analy- sis: A Tutorial and Survey. InProceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD). 6555... on Pithdoi:10.1145/3637528
- [73] Scott M Lundberg and Su-In Lee. 2017. A unified approach to interpreting model predictions.Advances in neural information processing systems30 (2017) on Pith
- [74] Takuya Maeda and Anabel Quan-Haase. 2024. When Human-AI Interactions Become Parasocial: Agency and Anthropomorphism in Affective Design. In Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Trans- parency(Rio de Janeiro, Brazil)(FAccT ’24). Association fo... on Pithdoi:10.1145/3630106.3658956
- [75] Wilson E Marcílio-Jr and Danilo M Eler. 2026. Navigating the Concept Space of Language Models.arXiv preprint arXiv:2603.23524(2026) on PitharXiv
- [76] Samuel Marks and Max Tegmark. 2023. The geometry of truth: Emergent linear structure in large language model representations of true/false datasets.arXiv preprint arXiv:2310.06824(2023) on PithPith's reviewarXiv
- [77] Jiahong Liu, Zexuan Qiu, Zhongyang Li, Quanyu Dai, Wenhao Yu, Jieming Zhu, Minda Hu, Menglin Yang, Tat-Seng Chua, and Irwin King. 2025. A Survey of Personalized Large Language Models: Progress and Future Directions.arXiv preprint arXiv:2502.11528(2025). https://arxiv.org/abs/2... on PitharXiv
- [78] Christina Lu, Jack Gallagher, Jonathan Michala, Kyle Fish, and Jack Lindsey on Pith
- [79] The Assistant Axis: Situating and Stabilizing the Default Persona of Language Models. arXiv:2601.10387 [cs.CL] https://arxiv.org/abs/2601.10387 on PitharXiv
- [80] Thomas McGrath, Andrei Kapishnikov, Nenad Tomašev, Adam Pearce, Martin Wattenberg, Demis Hassabis, Been Kim, Ulrich Paquet, and Vladimir Kramnik on Pith
- [81] Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov. 2022. Locating and Editing Factual Associations in GPT. InAdvances in Neural Information Processing Systems, Vol. 35. arXiv:2202.05262 [cs.CL] https://arxiv.org/abs/2202. 05262 Code and data at https://rome.baulab.info/ on PithPith's reviewarXiv
- [82] Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov. 2022. Locating and editing factual associations in gpt.Advances in neural information processing systems35 (2022), 17359–17372 on Pith
- [83] Tim Miller. 2023. Explainable AI is Dead, Long Live Explainable AI! Hypothesis- driven Decision Support using Evaluative AI. InProceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency (FAccT ’23). Association for Computing Machinery, New York, NY, U... on Pithdoi:10.1145/3593013
- [84] Samuel Marks, Johannes Treutlein, Trenton Bricken, Jack Lindsey, Jonathan Marcus, Siddharth Mishra-Sharma, Daniel Ziegler, Emmanuel Ameisen, Joshua Batson, Tim Belonax, Samuel R. Bowman, Shan Carter, Brian Chen, Hoagy Cunningham, Carson Denison, Florian Dietz, Satvik Golechha,... on PithPith's reviewarXiv
- [85] R. Thomas McCoy, Paul Soulos, Tal Linzen, and Paul Smolensky. 2026. The Emer- gent Symbolic Structure of Artificial Neural Networks. arXiv:2608.29530 [cs.CL] https://arxiv.org/abs/2608.29530 on PithPith's reviewarXiv
- [86] Russ McDonald. 2001.Shakespeare and the Arts of Language. Oxford University Press, Oxford on Pith
- [87] Hadas Orgad, Fazl Barez, Tal Haklay, Isabelle Lee, Marius Mosbach, Anja Reusch, Naomi Saphra, Byron C. Wallace, Sarah Wiegreffe, Eric Wong, Ian Tenney, and Mor Geva. 2026. Interpretability Can Be Actionable.arXiv preprint arXiv:2605.11161(2026). arXiv:2605.11161 [cs.LG] on PithPith's reviewarXiv
- [88] Nathan Paek, Yongyi Zang, Qihui Yang, and Randal Leistikow. 2025. Learning Interpretable Features in Audio Latent Spaces via Sparse Autoencoders.arXiv preprint arXiv:2510.23802(2025) on PitharXiv
- [89] Alexander Pan, Lijie Chen, and Jacob Steinhardt. 2024. LatentQA: Teach- ing LLMs to Decode Activations Into Natural Language.arXiv preprint arXiv:2412.08686(2024). https://arxiv.org/abs/2412.08686 on PitharXiv
- [90] Wenbo Pan, Zhichao Liu, Xianlong Wang, Haining Yu, and Xiaohua Jia. 2026. Towards Long-Horizon Interpretability: Efficient and Faithful Multi-Token At- tribution for Reasoning LLMs. InProceedings of the 43rd International Conference on Machine Learning (ICML) (PMLR, Vol. 306).... on PithPith's reviewarXiv
- [91] Jack Parker-Holder, Shlomi Fruchter, et al . 2025. Genie 3: A New Frontier for World Models. Google DeepMind. https://deepmind.google/discover/blog/ genie-3-a-new-frontier-for-world-models/ on Pith
- [92] Vince Miller, Mark J. Hill, and Tiago Moreira. 2026. Artificial intelligence and unconditional love: The rise of generative AI as an alternative form of mental health support.Communication and Change2 (2026), Article 14. doi:10.1007/ s44382-026-00036-1 Interviews + Reddit anal... on Pith
- [93] Neel Nanda, Andrew Lee, and Martin Wattenberg. 2023. Emergent linear representations in world models of self-supervised sequence models.arXiv preprint arXiv:2309.00941(2023) on PithPith's reviewarXiv
- [94] nostalgebraist. 2020. interpreting GPT: the logit lens — LessWrong — lesswrong.com. https://www.lesswrong.com/posts/AcKRB8wDpdaN6v6ru/ interpreting-gpt-the-logit-lens. [Accessed 12-09-2025] on Pith
- [95] Jason Phang, Michael Lampe, Lama Ahmad, Sandhini Agarwal, Cathy Mengying Fang, Auren R Liu, Valdemar Danry, Eunhae Lee, Samantha WT Chan, Pat Pataranutaporn, et al. 2025. Investigating affective use and emotional well- being on ChatGPT.arXiv preprint arXiv:2504.03888(2025) on PithPith's reviewarXiv
- [96] Daking Rai, Yilun Zhou, Shi Feng, Abulhair Saparov, and Ziyu Yao. 2024. A Prac- tical Review of Mechanistic Interpretability for Transformer-Based Language Models.arXiv preprint arXiv:2407.02646(2024). https://arxiv.org/abs/2407.02646 on PitharXiv
- [97] Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. 2016. " Why should I trust you?" Explaining the predictions of any classifier. InProceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. 1135–1144 on Pith
- [98] Nina Rimsky, Nick Gabrieli, Julian Schulz, Meg Tong, Evan Hubinger, and Alexander Turner. 2024. Steering llama 2 via contrastive activation addition. InProceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 15504–15522 on Pith
- [99] Tom Ritchey. 2018. General Morphological Analysis as a Basic Scientific Mod- elling Method.Technological Forecasting and Social Change126 (2018), 81–91. doi:10.1016/j.techfore.2017.05.027 on Pithdoi:10.1016/j.techfore.2017.05.027
- [100] Pat Pataranutaporn, Angela Vujic, David S. Kong, Pattie Maes, and Misha Sra on Pith
The first 100 of 156 references are listed here. All of them are in this paper's citation record.
Challenges and discussion
Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.
Come back by email
A short list of the week's newly reviewed papers. Nothing is sent until you confirm from your inbox, and every email has a link that stops it. No account.


No comments yet.
Sign in to comment