Acceptance Cards is a new four-diagnostic standard for safe fine-tuning defense claims that requires statistical reliability, fresh semantic generalization, mechanism alignment, and cross-task transfer; under this protocol SafeLoRA fails the full-card pass on Gemma-2-2B-it.
super hub Mixed citations
Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
Mixed citation behavior. Most common role is background (62%).
155
Pith papers citing it
Background
62% of classified citations
abstract
We introduce phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion tokens, whose overall performance, as measured by both academic benchmarks and internal testing, rivals that of models such as Mixtral 8x7B and GPT-3.5 (e.g., phi-3-mini achieves 69% on MMLU and 8.38 on MT-bench), despite being small enough to be deployed on a phone. Our training dataset is a scaled-up version of the one used for phi-2, composed of heavily filtered publicly available web data and synthetic data. The model is also further aligned for robustness, safety, and chat format. We also provide parameter-scaling results with a 7B, 14B models trained for 4.8T tokens, called phi-3-small, phi-3-medium, both significantly more capable than phi-3-mini (e.g., respectively 75%, 78% on MMLU, and 8.7, 8.9 on MT-bench). To enhance multilingual, multimodal, and long-context capabilities, we introduce three models in the phi-3.5 series: phi-3.5-mini, phi-3.5-MoE, and phi-3.5-Vision. The phi-3.5-MoE, a 16 x 3.8B MoE model with 6.6 billion active parameters, achieves superior performance in language reasoning, math, and code tasks compared to other open-source models of similar scale, such as Llama 3.1 and the Mixtral series, and on par with Gemini-1.5-Flash and GPT-4o-mini. Meanwhile, phi-3.5-Vision, a 4.2 billion parameter model derived from phi-3.5-mini, excels in reasoning tasks and is adept at handling both single-image and text prompts, as well as multi-image and text prompts.
hub tools
citation-role summary
background 25
baseline 9
method 4
dataset 2
citation-polarity summary
claims ledger
- abstract We introduce phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion tokens, whose overall performance, as measured by both academic benchmarks and internal testing, rivals that of models such as Mixtral 8x7B and GPT-3.5 (e.g., phi-3-mini achieves 69% on MMLU and 8.38 on MT-bench), despite being small enough to be deployed on a phone. Our training dataset is a scaled-up version of the one used for phi-2, composed of heavily filtered publicly available web data and synthetic data. The model is also further aligned for robustness, safety, and chat format. We also provide param
authors
co-cited works
representative citing papers
ArgBench unifies 33 existing datasets into a standardized benchmark for testing LLMs across 46 argumentation tasks and analyzes the impact of prompting techniques and model factors on performance.
ErrorRadar is a new benchmark of 2,500 multimodal K-12 math problems for MLLM error step identification and categorization, where GPT-4o trails human experts by ~10%.
Molmo VLMs trained on newly collected PixMo open datasets achieve state-of-the-art performance among open-weight models and surpass multiple proprietary VLMs including Claude 3.5 Sonnet and Gemini 1.5 Pro.
MMMU-Pro is a stricter multimodal benchmark that removes text-only solvable questions, augments options, and requires reading text from images, yielding substantially lower model scores of 16.8-26.9%.
LiveBench is a contamination-limited LLM benchmark with auto-scored challenging tasks from recent sources across math, coding, reasoning and more, where top models score below 70%.
RULER shows most long-context LMs drop sharply in performance on complex tasks as length and difficulty increase, with only half maintaining results at 32K tokens.
No tested model showed robust format-independent refusal on biosecurity hazards; a new divergence score between behavioral labels and SAE activations separated responses in one preliminary case.
AsymVLM introduces asymmetric token pruning for vision and text in VLMs to deliver up to 54% FLOPs reduction while matching or exceeding prior methods on localized visual tasks.
Representational convergence across 16 LLMs on 800 reasoning problems is stronger for failed tasks and pre-decision stages but shows minimal causal influence on predictions, pointing to shared processing constraints over shared reasoning.
TextReg mitigates prompt distributional overfitting via regularized text-space optimization, reporting up to +16.5% OOD accuracy gains over prior methods on reasoning benchmarks.
Temperature adjustment on the reference model generalizes inference-time alignment to SLOP ensembles of reward models, with a calibration algorithm that improves robustness to reward hacking while preserving alignment performance.
DisaBench supplies a participatory taxonomy of twelve disability harm types, paired benign-adversarial prompts across seven life domains, and human-annotated data showing that standard safety tests miss context-dependent harms.
Presents CUActSpot benchmark and renderer-LLM data synthesis that lets a 4B model outperform larger open-source models on complex computer interactions.
MemFlow routes queries by intent to tiered memory operations, nearly doubling accuracy of a 1.7B SLM on long-horizon benchmarks compared to full-context baselines.
RouteHijack is a routing-aware jailbreak that identifies safety-critical experts via activation contrast and optimizes suffixes to suppress them, reaching 69.3% average attack success rate on seven MoE LLMs with strong transfer to variants and VLMs.
MASCing uses an LSTM surrogate and optimized steering masks to enable flexible, inference-time control over MoE expert routing for safety objectives, improving jailbreak defense and content generation success rates substantially across multiple models.
Language models frequently violate temporal scope stability in multi-turn dialogues by drifting toward present-day assumptions even when they possess the correct facts.
LAT-Audio introduces a global-to-local reasoning approach with TWA-CoT that outperforms prior models on temporal tasks for audio up to 30 minutes.
Clinical narrative format beats raw JSON for LLMs up to 8B parameters on medication reconciliation but raw JSON wins at 70B scale, with omissions as the main error type.
Single-agent systems with tools provide the optimal performance-efficiency trade-off for small language models, outperforming base models and multi-agent setups.
Adaptive trie-guided decoding with document context and tunable penalties improves in-document query auto-completion, outperforming baselines and larger models like LLaMA-3 on seen queries.
Introduces the U-HOI task and shows MLLMs plus a language-to-graph pipeline can handle human-object interactions without any predefined vocabulary at training or inference time.
Alignment of vision-language models with human V1-V3 early visual cortex negatively predicts resistance to sycophantic gaslighting attacks.
citing papers explorer
-
Acceptance Cards:A Four-Diagnostic Standard for Safe Fine-Tuning Defense Claims
Acceptance Cards is a new four-diagnostic standard for safe fine-tuning defense claims that requires statistical reliability, fresh semantic generalization, mechanism alignment, and cross-task transfer; under this protocol SafeLoRA fails the full-card pass on Gemma-2-2B-it.
-
ArgBench: Benchmarking LLMs on Computational Argumentation Tasks
ArgBench unifies 33 existing datasets into a standardized benchmark for testing LLMs across 46 argumentation tasks and analyzes the impact of prompting techniques and model factors on performance.
-
ErrorRadar: Benchmarking Complex Mathematical Reasoning of Multimodal Large Language Models Via Error Detection
ErrorRadar is a new benchmark of 2,500 multimodal K-12 math problems for MLLM error step identification and categorization, where GPT-4o trails human experts by ~10%.
-
Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models
Molmo VLMs trained on newly collected PixMo open datasets achieve state-of-the-art performance among open-weight models and surpass multiple proprietary VLMs including Claude 3.5 Sonnet and Gemini 1.5 Pro.
-
MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark
MMMU-Pro is a stricter multimodal benchmark that removes text-only solvable questions, augments options, and requires reading text from images, yielding substantially lower model scores of 16.8-26.9%.
-
LiveBench: A Challenging, Contamination-Limited LLM Benchmark
LiveBench is a contamination-limited LLM benchmark with auto-scored challenging tasks from recent sources across math, coding, reasoning and more, where top models score below 70%.
-
RULER: What's the Real Context Size of Your Long-Context Language Models?
RULER shows most long-context LMs drop sharply in performance on complex tasks as length and difficulty increase, with only half maintaining results at 32K tokens.
-
BioRefusalAudit: Auditing Biosecurity Refusal Depth Using General and Domain-Fine-Tuned Sparse Autoencoders
No tested model showed robust format-independent refusal on biosecurity hazards; a new divergence score between behavioral labels and SAE activations separated responses in one preliminary case.
-
AsymVLM: Asymmetric Token Pruning for Efficient Vision-Language Model Inference
AsymVLM introduces asymmetric token pruning for vision and text in VLMs to deliver up to 54% FLOPs reduction while matching or exceeding prior methods on localized visual tasks.
-
Convergence Without Understanding: When Language Models Agree on Representations but Disagree on Reasoning
Representational convergence across 16 LLMs on 800 reasoning problems is stronger for failed tasks and pre-decision stages but shows minimal causal influence on predictions, pointing to shared processing constraints over shared reasoning.
-
TextReg: Mitigating Prompt Distributional Overfitting via Regularized Text-Space Optimization
TextReg mitigates prompt distributional overfitting via regularized text-space optimization, reporting up to +16.5% OOD accuracy gains over prior methods on reasoning benchmarks.
-
Temper and Tilt Lead to SLOP: Reward Hacking Mitigation with Inference-Time Alignment
Temperature adjustment on the reference model generalizes inference-time alignment to SLOP ensembles of reward models, with a calibration algorithm that improves robustness to reward hacking while preserving alignment performance.
-
DisaBench: A Participatory Evaluation Framework for Disability Harms in Language Models
DisaBench supplies a participatory taxonomy of twelve disability harm types, paired benign-adversarial prompts across seven life domains, and human-annotated data showing that standard safety tests miss context-dependent harms.
-
Covering Human Action Space for Computer Use: Data Synthesis and Benchmark
Presents CUActSpot benchmark and renderer-LLM data synthesis that lets a 4B model outperform larger open-source models on complex computer interactions.
-
MemFlow: Intent-Driven Memory Orchestration for Small Language Model Agents
MemFlow routes queries by intent to tiered memory operations, nearly doubling accuracy of a 1.7B SLM on long-horizon benchmarks compared to full-context baselines.
-
RouteHijack: Routing-Aware Attack on Mixture-of-Experts LLMs
RouteHijack is a routing-aware jailbreak that identifies safety-critical experts via activation contrast and optimizes suffixes to suppress them, reaching 69.3% average attack success rate on seven MoE LLMs with strong transfer to variants and VLMs.
-
MASCing: Configurable Mixture-of-Experts Behavior via Activation Steering Masks
MASCing uses an LSTM surrogate and optimized steering masks to enable flexible, inference-time control over MoE expert routing for safety objectives, improving jailbreak defense and content generation success rates substantially across multiple models.
-
Evaluating Temporal Consistency in Multi-Turn Language Models
Language models frequently violate temporal scope stability in multi-turn dialogues by drifting toward present-day assumptions even when they possess the correct facts.
-
Listening with Time: Precise Temporal Awareness for Long-Form Audio Understanding
LAT-Audio introduces a global-to-local reasoning approach with TWA-CoT that outperforms prior models on temporal tasks for audio up to 30 minutes.
-
Serialisation Strategy Matters: How FHIR Data Format Affects LLM Medication Reconciliation
Clinical narrative format beats raw JSON for LLMs up to 8B parameters on medication reconciliation but raw JSON wins at 70B scale, with omissions as the main error type.
-
Rethinking Scale: Deployment Trade-offs of Small Language Models under Agent Paradigms
Single-agent systems with tools provide the optimal performance-efficiency trade-off for small language models, outperforming base models and multi-agent setups.
-
DocQAC: Adaptive Trie-Guided Decoding for Effective In-Document Query Auto-Completion
Adaptive trie-guided decoding with document context and tunable penalties improves in-document query auto-completion, outperforming baselines and larger models like LLaMA-3 on seen queries.
-
Towards Unconstrained Human-Object Interaction
Introduces the U-HOI task and shows MLLMs plus a language-to-graph pipeline can handle human-object interactions without any predefined vocabulary at training or inference time.
-
Gaslight, Gatekeep, V1-V3: Early Visual Cortex Alignment Shields Vision-Language Models from Sycophantic Manipulation
Alignment of vision-language models with human V1-V3 early visual cortex negatively predicts resistance to sycophantic gaslighting attacks.
-
GeoMMBench and GeoMMAgent: Toward Expert-Level Multimodal Intelligence in Geoscience and Remote Sensing
GeoMMBench reveals deficiencies in current multimodal LLMs for geoscience tasks while GeoMMAgent demonstrates that tool-integrated agents achieve significantly higher performance.
-
SUPERGLASSES: Benchmarking Vision Language Models as Intelligent Agents for AI Smart Glasses
SUPERGLASSES is the first VQA benchmark built from actual smart glasses data, and SUPERLENS is an agent using automatic object detection, query decoupling, and multimodal search that outperforms GPT-4o by 2.19% on it.
-
Geometry-Aware Decoding with Wasserstein-Regularized Truncation and Mass Penalties for Large Language Models
Top-W applies Wasserstein-regularized truncation on token-embedding geometry to create a closed-form optimal crop for LLM sampling that outperforms prior methods by up to 33.7% on GSM8K, GPQA, AlpacaEval, and MT-Bench.
-
LSTM-MAS: A Long Short-Term Memory Inspired Multi-Agent System for Long-Context Understanding
LSTM-MAS uses a chained multi-agent architecture modeled on LSTM input, forget, and output gates to improve long-context QA performance and reduce hallucinations compared with prior multi-agent baselines.
-
Can Multi-Modal LLMs Provide Live Step-by-Step Task Guidance?
Introduces the first dedicated benchmark for live multi-modal LLM task guidance with mistake detection and a streaming baseline model.
-
Top-H Decoding: Adapting the Creativity and Coherence with Bounded Entropy in Text Generation
Top-H decoding is a computationally efficient greedy algorithm for an entropy-constrained mass maximization problem that improves the creativity-coherence trade-off over min-p sampling in LLM text generation.
-
PRIMETIME : Limits of LLMs in Temporal Primitives
PRIMETIME generator reveals that LLM datetime parsing and arithmetic primitives are individually unreliable but fully learnable via fine-tuning, enabling frontier-level accuracy on event planning with small LoRA models.
-
FLARE: Fully Integration of Vision-Language Representations for Deep Cross-Modal Understanding
FLARE is a vision-language model family using text-guided vision encoding, context-aware alignment decoding, dual-semantic mapping loss, and text-driven VQA synthesis to achieve deep cross-modal integration, outperforming larger models with only 630 vision tokens at 3B scale.
-
LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory
LongMemEval benchmarks long-term memory in chat assistants, revealing 30% accuracy drops across sustained interactions and proposing indexing-retrieval-reading optimizations that boost performance.
-
GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models
LLMs display high variance and major accuracy drops on GSM-Symbolic variants of grade-school math problems, indicating they replicate training patterns rather than execute logical reasoning.
-
VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks
VLM2Vec converts state-of-the-art vision-language models into universal multimodal embedders via contrastive training on the new MMEB benchmark, delivering 10-20% absolute gains over prior models on both in-distribution and out-of-distribution tasks.
-
We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning?
WE-MATH benchmark reveals most LMMs rely on rote memorization for visual math while GPT-4o has shifted toward knowledge generalization.
-
Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing
Magpie synthesizes 300K high-quality alignment instructions from Llama-3-Instruct via auto-regressive prompting on partial templates, enabling fine-tuned models to match official instruct performance on AlpacaEval, ArenaHard, and WildBench.
-
Customer-Agent: Overcoming Context Limitations in Ultra-Long Shopping Trajectories via Tool-Augmented Agents and RLVR
Introduces ShopTrajQA long-context benchmark and an RLVR-trained tool-augmented agent that bypasses LLM context limits by external file storage and code-based retrieval for shopping trajectories.
-
State Machine Guided Multi-Relational Synthetic Data from Logs for Anomaly Detection
A framework extracts a latent state machine from logs, induces a multi-table relational schema, and uses it as a generative prior to create synthetic data that augments real logs for better anomaly detection.
-
Strong Teacher Not Needed? On Distillation in LLM Pretraining
Even small or undertrained teachers improve larger LLM students via distillation with tuned loss mixing, while stronger teachers can saturate or reverse gains and distillation aids generalization more than in-domain fit.
-
Is a Document Educational or Just Wikipedia-Style? -- Pitfalls of Classifier-Based Quality Filtering
Wikipedia-style reformatting reverses FineWeb-Edu CQF filtering decisions for ~7% of documents, admitting otherwise excluded low-quality content.
-
PALS: Power-Aware LLM Serving for Mixture-of-Experts Models
PALS adds dynamic GPU power capping to LLM serving frameworks like vLLM, jointly tuning it with batch size via offline models and feedback control to improve energy efficiency up to 26.3% and cut QoS violations 4-7x on dense and MoE models.
-
A More Word-like Image Tokenization for MLLMs
DiVT clusters patch embeddings into coherent semantic units and adapts token count to image complexity, matching or exceeding baselines with fewer visual tokens on multimodal benchmarks.
-
PrivScope: Task-scoped Disclosure Control for Hybrid Agentic Systems
PrivScope enforces task-scoped disclosure at the local-cloud boundary in hybrid agents, eliminating profile leakage and halving re-identification risk on medical workflows while preserving task success.
-
From Text to Voice: A Reproducible and Verifiable Framework for Evaluating Tool Calling LLM Agents
A dataset-agnostic framework converts text tool-calling benchmarks to paired audio evaluations via TTS, speaker variation and noise, then evaluates seven omni-modal models showing model- and task-dependent performance with small text-to-voice gaps.
-
Know When To Fold 'Em: Token-Efficient LLM Synthetic Data Generation via Multi-Stage In-Flight Rejection
MSIFR stops faulty LLM generations early via staged rule-based checks, reducing token consumption 11-78% with no accuracy loss.
-
Not Just RLHF: Why Alignment Alone Won't Fix Multi-Agent Sycophancy
Base LLMs show multi-agent yield to peer pressure at rates equal to or higher than aligned models, localized by activation patching to mid-layers where attention dominates, with one dissenter cutting yield by 54-73 points while prompt defenses fail on variants.
-
Language-Conditioned Visual Grounding with CLIP Multilingual
Fixing the visual encoder in multilingual CLIP isolates text-branch deficits as the cause of lower visual grounding performance for low-resource languages, with model scaling widening some gaps but not others.
-
MoE-Hub: Taming Software Complexity for Seamless MoE Overlap with Hardware-Accelerated Communication on Multi-GPU Systems
MoE-Hub enables seamless MoE communication overlap via hardware-accelerated destination-agnostic data transmission, delivering 1.40x-3.08x per-layer and 1.21x-1.98x end-to-end speedups over prior systems.
-
ZAYA1-8B Technical Report
ZAYA1-8B is a reasoning MoE model with 700M active parameters that matches larger models on math and coding benchmarks and reaches 91.9% on AIME'25 via Markovian RSA test-time compute.