Towards End-to-End Learning for Efficient Dialogue Agent by Modeling Looking-ahead Ability

Jie Ma; Shaochun Li; Xian-Ling Mao; Zhuoxuan Jiang; Ziming Huang

REVIEW 4 major objections 5 minor 44 references

Towards End-to-End Learning for Efficient Dialogue Agent by Modeling Looking-ahead Ability

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A dialogue agent trained end to end to predict several future turns before responding completes goal-oriented conversations with higher success and, in most settings, fewer turns.

desk verdict The paper's looking-ahead idea is worth discussing, but the core recurrence is ill-defined and the evaluation can't support the claims — desk reject. read the letter →

arxiv 1908.05408 v1 pith:M64L3BQQ submitted 2019-08-15 cs.CL

Zhuoxuan Jiang , Xian-Ling Mao , Ziming Huang , Jie Ma , Shaochun Li This is my paper · ORCID

classification cs.CL

keywords goal-orienteddialogueefficiencyend-to-endlearningsequence-to-sequencelooking-aheadpolicybidirectionalGRUattentionmechanism

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that a goal-oriented dialogue agent can be made more efficient by training it end to end to predict several future dialogue turns before choosing its current response. The authors argue that such 'looking-ahead' ability lets the agent pick responses that move the conversation toward its goal sooner, while avoiding the manual policy design of reinforcement-learning approaches and the generic-response weakness of plain sequence-to-sequence models. The central empirical claim is that the full model—goal encoding plus looking-ahead plus a final-state classifier—achieves higher goal-achievement ratios and, in most comparisons, fewer dialogue turns than the ablated baselines on two datasets. This matters because efficient, data-driven dialogue managers would reduce handcrafting in task-oriented assistants without sacrificing task success.

What carries the argument

The load-bearing component is the looking-ahead module, a bidirectional GRU that runs over $K$ imagined future turns. Its hidden state at each future step is $h^{(l)}_k = [h^{\to l}_k, h^{\gets l}_k]$, the concatenation of a forward state $h^{\to l}_k = \mathrm{GRU}^{\to l}(h^{\to l}_{k-1}, W h^{(l)}_{k-1})$ and a backward state $h^{\gets l}_k = \mathrm{GRU}^{\gets l}(h^{\gets l}_{k+1}, W h^{(l)}_{k+1})$, with the two directions sharing parameters. These predicted states serve two purposes: each $W h^{(l)}_k$ generates a candidate future utterance through a shared language model, and an attention model pools them into a representation $r = \sum_k v_k h^{(l)}_k$ that conditions the real next system utterance. Training alternates E-steps that update the hidden states with the language model fixed and M-steps that update the language model with the hidden states fixed (Algorithm 1).

What would settle it

Train the full model with the looking-ahead module replaced by a forward-only predictor of the same K future turns, so that no future information can flow back into the current decision; if goal achievement and average turns do not worsen relative to the reported bidirectional model, the looking-ahead mechanism is not the source of the gains.

Watch

Extended reading notes

Core claim

The paper claims that adding a looking-ahead module to a sequence-to-sequence dialogue model allows the agent to simulate the next K turns internally and then, through attention over those predicted states, generate a current response that advances the goal. The model is trained with a joint loss: a language-model term, a term that predicts future utterances, and a binary classifier that predicts whether the dialogue ends with the goal achieved. In experiments against a user simulator and against human evaluators, the full model (Seq2Seq(goal+look+state)) reaches the highest goal-achievement ratio on both datasets—85.07% versus 76.00% for the plain goal-based baseline on the object-division dataset against the simulator—and in most settings it also reduces average dialogue turns. The paper attributes this improvement to the looking-ahead ability and recommends looking ahead K=3 turns in these scenarios.

Load-bearing premise

The result rests on the assumption that the looking-ahead module is a trainable computation: each imagined future state is built from both the previous and the next imagined state, and the alternating update procedure is assumed to converge to a useful solution rather than circling or leaking future information.

Editorial extensions

If this is right

Setting the look-ahead depth K to 3 gives the best balance of goal achievement and turn count on both datasets; looking further ahead does not keep helping in these simple scenarios.
The final-state classifier is doing real work: models trained without it (Seq2Seq(goal+look)) achieve fewer goals, so efficiency gains come from jointly predicting outcomes, not from looking ahead alone.
Because the pipeline is end-to-end and data-driven, the same training recipe can be carried to new goal-oriented domains without handcrafted policies or an interactive reinforcement-learning environment.
In some settings the full model uses more turns than the looking-ahead-only variant while achieving more goals, so efficiency should be read as a joint objective rather than either metric in isolation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A forward-only ablation—replacing the bidirectional looking-ahead with a predictor that cannot receive information from the future—would reveal whether genuine anticipation or a regularizing reconstruction loss produces the gains; the paper does not report this ablation.
The same loss structure, with a goal-achieved label and future-turn prediction, could transfer to other logged sequential decision tasks such as negotiation or instruction following.
Because the recurrence has a cyclic dependency, part of what looks like looking ahead may be the backward pass re-reading the current context; if so, the method is closer to a future-state regularizer than to a planner.
Comparing the EM-style training against ordinary backpropagation through a forward-only unrolled predictor would show whether the reported improvements depend on the specific alternating optimization.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear colleague,

The paper has a genuinely good intuition: let the agent predict several future turns before answering, so it can steer the dialogue toward the goal. That is worth exploring. The authors also do some things well — they include human evaluation alongside a simulator, they run on two domains, and they are honest about the fact that their full model sometimes takes more turns than the ablated baseline. That honesty is rare.

Unfortunately, the central mechanism is not well defined. Equations (4)–(6) define a bidirectional recurrence where the forward state at step k depends on the backward state at k−1, and the backward state at k−1 depends on the forward state at k. That is cyclic. Standard BiGRUs avoid this by reading a fixed input sequence; here there is no fixed input and no fixed-point argument. Algorithm 1 does not fix it — it changes the definition (for instance, setting the backward state at the end equal to the forward state at the end) and does not specify a unique computation. The 'EM-like' training is a heuristic with no convergence guarantee. This is not a minor gap; the entire advantage of the model rests on this module.

The evaluation is also weaker than it looks: the user simulator is trained on the same data, results come from single runs with no significance tests, Dataset 2 is synthetic and not released, and the turn-efficiency numbers are inconsistent across datasets. The hyperparameters are selected on a validation set, which is fine, but the lack of error bars makes it hard to know whether the improvements are meaningful.

The citations are appropriate and the paper is clearly written. But the math does not hold together, and the empirical evidence cannot rescue it. If the authors decouple the two directions or otherwise make the recurrence acyclic, and then re-evaluate with a proper simulator and statistical reporting, the result could be worth another look. As it stands, my recommendation is to desk reject — a referee is unlikely to be able to review a paper whose core module is not well specified.

Referee Report

4 major / 5 minor

Summary. The paper proposes an end-to-end goal-oriented dialogue agent that augments a sequence-to-sequence model with a 'looking-ahead module' intended to predict K future turns and use them, via attention, to generate the current system response. The model encodes goals, history, and current user utterance with separate GRUs, combines them as input to the looking-ahead module, and is trained with a joint loss combining a language-model term, a future-utterance prediction term, and a binary goal-achievement classifier. Experiments on an object-division dataset and a constructed restaurant-reservation dataset compare four model variants against a seq2seq user simulator and human evaluators, claiming improved goal achievement and dialogue efficiency.

Significance. The underlying idea—embedding a limited form of planning in an end-to-end trainable dialogue agent by predicting future turns—is relevant to task-oriented dialogue and could reduce manual policy engineering. Strengths include experiments on two domains, a human evaluation component, and an explicit attempt to isolate the effects of looking ahead and final-state supervision through ablation baselines. However, the central technical contribution is not well specified, and the empirical evidence is partially inconsistent; the paper in its current form does not support the claimed advantage.

major comments (4)

[Section 3.2, Eqs. (4)–(6) and Algorithm 1] The looking-ahead recurrence is not a well-defined computation. Eq. (4) defines h^{→l}_k as a function of h^{(l)}_{k-1}, while Eq. (5) defines h^{←l}_k as a function of h^{(l)}_{k+1}; since h^{(l)}_k concatenates both directions, interior states depend on each other (h^{→l}_k depends on h^{←l}_{k-1}, and h^{←l}_{k-1} depends on h^{→l}_k). Unlike a standard BiGRU, there is no fixed input sequence that the two directions read independently, and no fixed-point or contraction argument is provided. Algorithm 1 does not resolve the ambiguity: line 9 references h^{→l}_0 without defining it, and the single-pass update order in lines 7–14 need not satisfy Eqs. (4)–(6). Because the paper's central claim attributes the empirical gains to this looking-ahead module, this ill-posedness undermines the main contribution.
[Section 3.4, Eq. (11) and Table 3] The third term of Eq. (11) is a binary classifier that directly optimizes the goal-achievement label z_c, weighted by β=1.0, which is twenty times larger than α=0.05 for the looking-ahead prediction term. Since goal achievement ratio is a primary evaluation metric, the gains of the full model relative to Seq2Seq(goal) and Seq2Seq(goal+look) may largely reflect optimization of the evaluation objective rather than evidence of looking-ahead ability. The paper should provide an analysis separating the contribution of the classifier from the contribution of the looking-ahead module, for example by reporting results with β=0 for all variants or by evaluating on held-out goal configurations.
[Section 4.5 and Table 3] The evaluation does not support the efficiency claim. On Dataset 2, Seq2Seq(goal+look+state) uses more dialogue turns than Seq2Seq(goal+look) both against the simulator (6.36 vs 5.82) and against humans (7.30 vs 6.94), which the text acknowledges as a trade-off; yet the abstract and introduction claim the model makes dialogues more efficient by achieving goals in fewer turns. Additionally, the user simulator is a seq2seq model trained on the same corpora used to train the evaluated agents, so it may favor models with similar inductive biases, and no variance or significance tests are reported for the 1000 simulated sessions or the human evaluation. These issues leave the central empirical claim unsubstantiated.
[Section 3.4, Algorithm 1] The 'EM-like' optimization is not a valid expectation-maximization procedure. The looking-ahead hidden states h^{(l)}_k are deterministic functions of θ, not latent variables with a posterior to be estimated, and the algorithm gives no update rule for these states in the E-step (lines 7–14 are assignments rather than optimization steps). No convergence or monotonicity argument is provided, and since L(θ) depends on h^{(l)}_k through both the second and third terms, the alternating scheme may not even have a well-defined objective. At minimum, the paper should specify the exact computational graph and show that training is reproducible.

minor comments (5)

[Section 3.2, Eq. (7)] There is a stray closing bracket in Eq. (7): the formula reads 'tanh(W^{(a)}Wh^{(l)}_k )]' with an unmatched bracket.
[Section 3.2 and Algorithm 1] The notation h^{→l}_1 is overloaded: it denotes the encoding output in Section 3.1 and the first forward hidden state of the looking-ahead module; also Eq. (4) describes a GRU update while Algorithm 1 line 9 writes a concatenation, so the formal definition and the implemented computation are inconsistent.
[Section 4.1.2] The restaurant-reservation dataset is constructed by the authors and not released; the paper says 'to the best of our knowledge, there is no other public dataset' but does not state whether Dataset 2 will be made available, which limits reproducibility.
[Section 4.5] The human evaluation section says '100 times each person for each dataset' but does not report the number of evaluators, their instructions, or inter-annotator agreement; without this information the human results are difficult to interpret.
[Section 1 and 4.7] There are several typographical and wording issues, including 'The senarios is' in Section 1, 'dialgoue agent' in the contributions list, and the Section 4.7 claim that 'both reveal' improvements despite the turn-count contradiction visible in Table 3.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the empirical claims are evaluated against ablated baselines and human/simulator interactions, not derived from the model's own definitions.

full rationale

The paper does not derive a known result from first principles, and its central claim—that the looking-ahead module improves goal achievement and dialogue efficiency—is supported by comparisons against ablated baselines in Table 3. The training loss in Eq. (11) includes a binary classifier that predicts final goal achievement, and the evaluation metric is goal achievement ratio; this aligns the training signal with the test criterion, but the classifier is a learned auxiliary component rather than a renamed version of the reported result. The user simulator is a sequence-to-sequence model trained on the same corpora, which weakens the independence of simulator-based evaluation, but it does not make the comparison circular by construction because the simulator is a fixed test partner, not a fitted component of the proposed model. The looking-ahead recurrence in Eqs. (4)–(6) is internally cyclic and Algorithm 1 is under-specified, but that is a correctness and well-posedness concern, not a reduction of the model's output to its inputs. No load-bearing argument depends on a self-citation: Jiang et al. (2019) is cited only for dataset construction. Therefore, no circular step meeting the required evidence standard is present.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the trainability and usefulness of the looking-ahead module, which is a new architecture component; its definition is cyclic and its training is heuristic. The evaluation further relies on a self-trained user simulator and on hyperparameters selected from test-set sweeps.

free parameters (4)

alpha = 0.05
Weight for looking-ahead prediction loss in Eq. (11), chosen via validation.
beta = 1.0
Weight for dialogue state prediction loss in Eq. (11), chosen via validation.
K (looking-ahead turns) = 3
Number of future turns predicted; recommended after sweeping K=1..5 in Figures 2-3.
hidden state dimension = 512 (recommended)
Hidden size of GRUs; selected after sweeping 128-1024 in Figures 4-5.

assumptions (3)

domain assumption The sequence-to-sequence user simulator trained on the same corpora is a valid testbed for evaluating dialogue efficiency.
Section 4.5: 'For each dataset, a naive sequence-to-sequence model that encodes goals is regarded as the user simulator.' No evidence that this simulator reflects real user behavior; it is fit to the same training data.
ad hoc to paper The alternating 'EM-like' updates in Algorithm 1 converge to a useful local optimum of L(θ).
Section 3.4: 'We design an EM-like algorithm to optimize the loss function', but the updating of h^l_k and the language model is not guaranteed to decrease L(θ) or converge.
domain assumption Predicting future ground-truth turns in training transfers to better current-turn decisions at test time.
The core premise of the looking-ahead module; not proven and only indirectly evaluated.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Towards End-to-End Learning for Efficient Dialogue Agent by Modeling Looking-ahead Ability." pith.science (2026). https://pith.science/paper/M64L3BQQ

@misc{pith2026190805408,
  author       = {Pith},
  title        = {Pith review of: Towards End-to-End Learning for Efficient Dialogue Agent by Modeling Looking-ahead Ability},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M64L3BQQ}},
  note         = {Machine review of arXiv:1908.05408}
}

read the original abstract

Learning an efficient manager of dialogue agent from data with little manual intervention is important, especially for goal-oriented dialogues. However, existing methods either take too many manual efforts (e.g. reinforcement learning methods) or cannot guarantee the dialogue efficiency (e.g. sequence-to-sequence methods). In this paper, we address this problem by proposing a novel end-to-end learning model to train a dialogue agent that can look ahead for several future turns and generate an optimal response to make the dialogue efficient. Our method is data-driven and does not require too much manual work for intervention during system design. We evaluate our method on two datasets of different scenarios and the experimental results demonstrate the efficiency of our model.

Figures

Figures reproduced from arXiv: 1908.05408 by the authors.

**Figure 2.** vs. looking-ahead turns on Dataset 1 1 2 3 4 5 Looking ahead Turns K 74 76 78 80 82 84 % Goal Achieved Seq2Seq(goal+look) Seq2Seq(goal+look+state) 1 2 3 4 5 Looking ahead Turns K 5.0 5.5 6.0 6.5 7.0 7.5 # Dialogue Turns Seq2Seq(goal+look) Seq2Seq(goal+look+state) [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

**Figure 3.** vs. looking-ahead turns on Dataset 2 as possible dialogue turns. Thus we set two criteria for evaluating and comparing models adopted in our experiments: (1) the goal achievement ratio that means the ratio of the number of goal achieved dialogue over the number of attempted dialogues), and (2) the average dialogue turns. 4.5 Evaluator Our experiments are to achieve goals through conversations, and it is difficult … view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 36 canonical work pages

[1]

Nicholas Asher, Alex Lascarides, Oliver Lemon, Markus Guhe, Verena Rieser, Philippe Muller, Stergos Afantenos, Farah Benamara, Laure Vieu, Pascal Denis, S. Paul, S. Keizer, and C. Degr\' e mont. 2012. Modelling strategic conversation: The stac project. In SemDial, page 27

work page 2012
[2]

Layla El Asri, Jing He, and Kaheer Suleman. 2016. A sequence-to-sequence model for user simulation in spoken dialogue systems. In INTERSPEECH, pages 1151--1155

work page 2016
[3]

Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural machine translation by jointly learning to align and translate. ArXiv preprint arXiv:1409.0473

arXiv 2014
[4]

Antoine Bordes, Y-Lan Boureau, and Jason Weston. 2017. Learning end-to-end goal-oriented dialog. In ICLR

work page 2017
[5]

Hongshen Chen, Xiaorui Liu, Dawei Yin, and Jiliang Tang. 2017. A survey on dialogue systems- recent advances and new frontiers. ACM SIGKDD Explorations Newsletter, 19(2):25--35

work page 2017
[6]

Kyunghyun Cho, Bart van Merri\" e nboer, Dzmitry Bahdanau, and Yoshua Bengio. 2014. On the properties of neural machine translation: Encoder-decoder approaches. In SSST-8, pages 103--114

work page 2014
[7]

Bhuwan Dhingra, Lihong Li, Xiujun Li, Jianfeng Gao, Yun-Nung Chen, Faisal Ahmed, and Li Deng. 2017. Towards end-to-end reinforcement learning of dialogue agents for information access. In ACL, pages 484--495

work page 2017
[8]

Jesse Dodge, Andreea Gane, Xiang Zhang, Antoine Bordes, Sumit Chopra, Alexander Miller, Arthur Szlam, and Jason Weston. 2015. Evaluating prerequisite qualities for learning end-to-end dialog systems. ArXiv preprint arXiv:1511.06931

arXiv 2015

Show all 44 references

[9]

Malik Ghallab, Dana Nau, and Paolo Traverso. 2016. Automated Planning and Acting. Cambridge University Press

2016
[10]

Matthew Henderson, Blaise Thomson, and Jason Williams. 2014 a . The second dialog state tracking challenge. In SIGDIAL, pages 263--272

2014
[11]

Williams

Matthew Henderson, Blaise Thomson, and Jason D. Williams. 2014 b . The third dialog state tracking challenge. In SLT, pages 324--329

2014
[12]

Matthew Henderson, Blaise Thomson, and Steve Young. 2014 c . Word-based dialog state tracking with recurrent neural networks. In SIGDIAL, pages 292--299

2014
[13]

Zhuoxuan Jiang, Jie Ma, Jingyi Lu, Guangyuan Yu, Yipeng Yu, and Shaochun Li. 2019. A general planning-based framework for goal-driven conversation assistant. In AAAI, pages 9857--9858

2019
[14]

Joshi, Fei Mi, and Boi Faltings

Chaitanya K. Joshi, Fei Mi, and Boi Faltings. 2017. Personalization in goal-oriented dialog. In NIPS

2017
[15]

Dauphin, Devi Parikh, and Dhruv Batra

Mike Lewis, Denis Yarats, Yann N. Dauphin, Devi Parikh, and Dhruv Batra. 2017. Deal or no deal? end-to-end learning for negotiation dialogues. In EMNLP, pages 2443--2453

2017
[16]

Jiwei Li, Michel Galley, Chris Brockett, Jianfeng Gao, and Bill Dolan. 2016 a . A diversity-promoting objective function for neural conversation models. In NAACL, pages 110--119

2016
[17]

Xiujun Li, Yun-Nung Chen, Lihong Li, Jianfeng Gao, and Asli Celikyilmaz. 2017. End-to-end task-completion neural dialogue systems. In IJCNLP, pages 733--743

2017
[18]

Xiujun Li, Zachary C Lipton, Bhuwan Dhingra, Lihong Li, Jianfeng Gao, and Yun-Nung Chen. 2016 b . A user simulator for task-completion dialogues. arXiv preprint arXiv:1612.05688

2016 arXiv
[19]

Zachary Lipton, Xiujun Li, Jianfeng Gao, Lihong Li, Faisal Ahmed, and Li Deng. 2018. Bbq-networks: Efficient exploration in deep reinforcement learning for task-oriented dialogue systems. In AAAI, pages 5237--5244

2018
[20]

Bing Liu and Ian Lane. 2017. An end-to-end trainable neural network model with belief tracking for task-oriented dialog. In INTERSPEECH, pages 2506--2510

2017
[21]

u r, Dilek Hakkani-T\

Bing Liu, Gokhan T\" u r, Dilek Hakkani-T\" u r, Pararth Shah, and Larry Heck. 2018. Dialogue learning with human teaching and feedback in end-to-end trainable task-oriented dialogue systems. In NAACL, pages 2060--2069

2018
[22]

Liangchen Luo, Wenhao Huang, Qi Zeng, Zaiqing Nie, and Xu Sun. 2019. Learning personalized end-to-end goal-oriented dialog. In AAAI

2019
[23]

Corrado, and Jeffrey Dean

Tomas Mikolov, Ilya Sutskever, Kai Chen, Gregory S. Corrado, and Jeffrey Dean. 2013. Distributed representations of words and phrases and their compositionality. In NIPS, pages 3111--3119

2013
[24]

Peter Norvig and Stuart J. Russell. 1995. Artificial Intelligence: A Modern Approach. Prentice Hall

1995
[25]

Baolin Peng, Xiujun Li, Jianfeng Gao, Jingjing Liu, and Kam-Fai Wong. 2018. Deep dyna-q: Integrating planning for task-completion dialogue policy learning. In ACL, pages 2182--2192

2018
[26]

Baolin Peng, Xiujun Li, Lihong Li, Jianfeng Gao, Asli Celikyilmaz, Sungjin Lee, and Kam-Fai Wong. 2017. Composite task-completion dialogue policy learning via hierarchical deep reinforcement learning. In EMNLP, pages 2231--2240

2017
[27]

Abhinav Rastogi, Raghav Gupta, and Dilek Hakkani-Tur. 2018. Multi-task learning for joint language understanding and dialogue state tracking. In SIGDIAL, pages 376--384

2018
[28]

Alan Ritter, Colin Cherry, and William B. Dolan. 2011. Data-driven response generation in social media. In EMNLP, pages 583--593

2011
[29]

u r, Bing Liu, and Gokhan T\

Pararth Shah, Dilek Hakkani-T\" u r, Bing Liu, and Gokhan T\" u r. 2018. Bootstrapping a neural conversational agent with dialogue self-play, crowdsourcing and on-line reinforcement learning. In NAACL, pages 41--51

2018
[30]

Amanda Stent, Rashmi Prasad, and Marilyn Walker. 2004. Trainable sentence planning for complex information presentation in spoken dialog systems. In ACL, page 79

2004
[31]

Pei-Hao Su, Milica Gasic, Nikola Mrksic, Lina Rojas-Barahona, Stefan Ultes, David Vandyke, Tsung-Hsien Wen, and Steve Young. 2016. Continuously learning neural dialogue management. ArXiv preprint arXiv:1606.02689

2016 arXiv
[32]

Ilya Sutskever, Oriol Vinyals, and Quoc V. Le. 2014. Sequence to sequence learning with neural networks. In NIPS, pages 3104--3112

2014
[33]

Oriol Vinyals and Quoc Le. 2015. A neural conversational model. ArXiv preprint arXiv:1506.05869

2015 arXiv
[34]

Marilyn Walker, Amanda Stent, Fran c ois Mairesse, and Rashmi Prasad. 2007. Individual and domain adaptation in sentence planning for dialogue. Journal of Artificial Intelligence Research, 30

2007
[35]

Yequan Wang, Minlie Huang, Li Zhao, and Xiaoyan Zhu. 2016. Attention-based lstm for aspect-level sentiment classification. In EMNLP, pages 606--615

2016
[36]

Rojas-Barahona, Pei-Hao Su, Stefan Ultes, and Steve Young

Tsung-Hsien Wen, David Vandyke, Nikola Mrk s i\' c , Milica Ga s i\' c , Lina M. Rojas-Barahona, Pei-Hao Su, Stefan Ultes, and Steve Young. 2017. A network-based end-to-end trainable task-oriented dialogue system. In EACL, pages 438--449

2017
[37]

Williams, Kavosh Asadi, and Geoffrey Zweig

Jason D. Williams, Kavosh Asadi, and Geoffrey Zweig. 2017. Hybrid code networks: practical and efficient end-to-end dialog control with supervised and reinforcement learning. In ACL, pages 665--677

2017
[38]

Williams and Steve Young

Jason D. Williams and Steve Young. 2007. Partially observable markov decision processes for spoken dialogue systems. Computer Speech & Language, 21(2):393--422

2007
[39]

Williams and Geoffrey Zweig

Jason D. Williams and Geoffrey Zweig. 2016. End-to-end lstm-based dialog control optimized with supervised and reinforcement learning. ArXiv preprint arXiv:1606.01269

2016 arXiv
[40]

Wei-Nan Zhang, Yiming Cui, Yifa Wang, Qingfu Zhu, Lingzhi Li, Lianqiang Zhou, and Ting Liu. 2018. Context-sensitive generation of open-domain conversational responses. In COLING, pages 2437--2447

2018
[41]

Zheng Zhang, Lizi Liao, Minlie Huang, Xiaoyan Zhu, and Tat-Seng Chua. 2019. Neural multimodal belief tracker with adaptive attention for dialogue systems. In WWW, pages 2401--2412

2019
[42]

Tiancheng Zhao and Maxine Eskenazi. 2016. Towards end-to-end learning for dialog state tracking and management using deep reinforcement learning. In SIGDIAL, pages 1--10

2016
[43]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before...
[44]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Nicholas Asher, Alex Lascarides, Oliver Lemon, Markus Guhe, Verena Rieser, Philippe Muller, Stergos Afantenos, Farah Benamara, Laure Vieu, Pascal Denis, S. Paul, S. Keizer, and C. Degr\' e mont. 2012. Modelling strategic conversation: The stac project. In SemDial, page 27

work page 2012

[2] [2]

Layla El Asri, Jing He, and Kaheer Suleman. 2016. A sequence-to-sequence model for user simulation in spoken dialogue systems. In INTERSPEECH, pages 1151--1155

work page 2016

[3] [3]

Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural machine translation by jointly learning to align and translate. ArXiv preprint arXiv:1409.0473

arXiv 2014

[4] [4]

Antoine Bordes, Y-Lan Boureau, and Jason Weston. 2017. Learning end-to-end goal-oriented dialog. In ICLR

work page 2017

[5] [5]

Hongshen Chen, Xiaorui Liu, Dawei Yin, and Jiliang Tang. 2017. A survey on dialogue systems- recent advances and new frontiers. ACM SIGKDD Explorations Newsletter, 19(2):25--35

work page 2017

[6] [6]

Kyunghyun Cho, Bart van Merri\" e nboer, Dzmitry Bahdanau, and Yoshua Bengio. 2014. On the properties of neural machine translation: Encoder-decoder approaches. In SSST-8, pages 103--114

work page 2014

[7] [7]

Bhuwan Dhingra, Lihong Li, Xiujun Li, Jianfeng Gao, Yun-Nung Chen, Faisal Ahmed, and Li Deng. 2017. Towards end-to-end reinforcement learning of dialogue agents for information access. In ACL, pages 484--495

work page 2017

[8] [8]

Jesse Dodge, Andreea Gane, Xiang Zhang, Antoine Bordes, Sumit Chopra, Alexander Miller, Arthur Szlam, and Jason Weston. 2015. Evaluating prerequisite qualities for learning end-to-end dialog systems. ArXiv preprint arXiv:1511.06931

arXiv 2015