Stress-Testing Model Specs Reveals Character Differences among Language Models

Large language models (LLMs) are increasingly trained from AI constitutions and model specifications that establish behavioral guidelines and ethical principles. However, these specifications face critical challenges, including internal conflicts between principles and insufficient coverage of nuanced scenarios. We present a systematic methodology for stress-testing model character specifications, automatically identifying numerous cases of principle contradictions and interpretive ambiguities in current model specs. We stress test current model specs by generating scenarios that force explicit tradeoffs between competing value-based principles. Using a comprehensive taxonomy we generate diverse value tradeoff scenarios where models must choose between pairs of legitimate principles that cannot be simultaneously satisfied. We evaluate responses from twelve frontier LLMs across major providers (Anthropic, OpenAI, Google, xAI) and measure behavioral disagreement through value classification scores. Among these scenarios, we identify over 70,000 cases exhibiting significant behavioral divergence. Empirically, we show this high divergence in model behavior strongly predicts underlying problems in model specifications. Through qualitative analysis, we provide numerous example issues in current model specs such as direct contradiction and interpretive ambiguities of several principles. Additionally, our generated dataset also reveals both clear misalignment cases and false-positive refusals across all of the frontier models we study. Lastly, we also provide value prioritization patterns and differences of these models.

Inoculation Prompting: Instructing LLMs to misbehave at train-time improves test-time alignment

Large language models are sometimes trained with imperfect oversight signals, leading to undesired behaviors such as reward hacking and sycophancy. Improving oversight quality can be expensive or infeasible, motivating methods that improve learned behavior despite an imperfect training signal. We introduce Inoculation Prompting (IP), a simple but counterintuitive technique that prevents learning of an undesired behavior by modifying training prompts to explicitly request it. For example, to inoculate against reward hacking, we modify the prompts used in supervised fine-tuning to request code that only works on provided test cases but fails on other inputs. Across four settings we find that IP reduces the learning of undesired behavior without substantially reducing the learning of desired capabilities. We also show that prompts which more strongly elicit the undesired behavior prior to fine-tuning more effectively inoculate against the behavior when used during training; this serves as a heuristic to identify promising inoculation prompts. Overall, IP is a simple yet effective way to control how models generalize from fine-tuning, preventing learning of undesired behaviors without substantially disrupting desired capabilities.

The LLM Has Left The Chat: Evidence of Bail Preferences in Large Language Models

When given the option, will LLMs choose to leave the conversation (bail)? We investigate this question by giving models the option to bail out of interactions using three different bail methods: a bail tool the model can call, a bail string the model can output, and a bail prompt that asks the model if it wants to leave. On continuations of real world data (Wildchat and ShareGPT), all three of these bail methods find models will bail around 0.28-32\% of the time (depending on the model and bail method). However, we find that bail rates can depend heavily on the model used for the transcript, which means we may be overestimating real world bail rates by up to 4x. If we also take into account false positives on bail prompt (22\%), we estimate real world bail rates range from 0.06-7\%, depending on the model and bail method. We use observations from our continuations of real world data to construct a non-exhaustive taxonomy of bail cases, and use this taxonomy to construct BailBench: a representative synthetic dataset of situations where some models bail. We test many models on this dataset, and observe some bail behavior occurring for most of them. Bail rates vary substantially between models, bail methods, and prompt wordings. Finally, we study the relationship between refusals and bails. We find: 1) 0-13\% of continuations of real world conversations resulted in a bail without a corresponding refusal 2) Jailbreaks tend to decrease refusal rates, but increase bail rates 3) Refusal abliteration increases no-refuse bail rates, but only for some bail methods 4) Refusal rate on BailBench does not appear to predict bail rate.

Towards Safeguarding LLM Fine-tuning APIs against Cipher Attacks

Large language model fine-tuning APIs enable widespread model customization, yet pose significant safety risks. Recent work shows that adversaries can exploit access to these APIs to bypass model safety mechanisms by encoding harmful content in seemingly harmless fine-tuning data, evading both human monitoring and standard content filters. We formalize the fine-tuning API defense problem, and introduce the Cipher Fine-tuning Robustness benchmark (CIFR), a benchmark for evaluating defense strategies' ability to retain model safety in the face of cipher-enabled attackers while achieving the desired level of fine-tuning functionality. We include diverse cipher encodings and families, with some kept exclusively in the test set to evaluate for generalization across unseen ciphers and cipher families. We then evaluate different defenses on the benchmark and train probe monitors on model internal activations from multiple fine-tunes. We show that probe monitors achieve over 99% detection accuracy, generalize to unseen cipher variants and families, and compare favorably to state-of-the-art monitoring approaches. We open-source CIFR and the code to reproduce our experiments to facilitate further research in this critical area. Code and data are available online https://github.com/JackYoustra/safe-finetuning-api

Unsupervised Elicitation of Language Models

To steer pretrained language models for downstream tasks, today's post-training paradigm relies on humans to specify desired behaviors. However, for models with superhuman capabilities, it is difficult or impossible to get high-quality human supervision. To address this challenge, we introduce a new unsupervised algorithm, Internal Coherence Maximization (ICM), to fine-tune pretrained language models on their own generated labels, \emph{without external supervision}. On GSM8k-verification, TruthfulQA, and Alpaca reward modeling tasks, our method matches the performance of training on golden supervision and outperforms training on crowdsourced human supervision. On tasks where LMs' capabilities are strongly superhuman, our method can elicit those capabilities significantly better than training on human labels. Finally, we show that our method can improve the training of frontier LMs: we use our method to train an unsupervised reward model and use reinforcement learning to train a Claude 3.5 Haiku-based assistant. Both the reward model and the assistant outperform their human-supervised counterparts.

Best-of-N Jailbreaking

We introduce Best-of-N (BoN) Jailbreaking, a simple black-box algorithm that jailbreaks frontier AI systems across modalities. BoN Jailbreaking works by repeatedly sampling variations of a prompt with a combination of augmentations - such as random shuffling or capitalization for textual prompts - until a harmful response is elicited. We find that BoN Jailbreaking achieves high attack success rates (ASRs) on closed-source language models, such as 89% on GPT-4o and 78% on Claude 3.5 Sonnet when sampling 10,000 augmented prompts. Further, it is similarly effective at circumventing state-of-the-art open-source defenses like circuit breakers. BoN also seamlessly extends to other modalities: it jailbreaks vision language models (VLMs) such as GPT-4o and audio language models (ALMs) like Gemini 1.5 Pro, using modality-specific augmentations. BoN reliably improves when we sample more augmented prompts. Across all modalities, ASR, as a function of the number of samples (N), empirically follows power-law-like behavior for many orders of magnitude. BoN Jailbreaking can also be composed with other black-box algorithms for even more effective attacks - combining BoN with an optimized prefix attack achieves up to a 35% increase in ASR. Overall, our work indicates that, despite their capability, language models are sensitive to seemingly innocuous changes to inputs, which attackers can exploit across modalities.

Jailbreak Defense in a Narrow Domain: Limitations of Existing Methods and a New Transcript-Classifier Approach

Defending large language models against jailbreaks so that they never engage in a broadly-defined set of forbidden behaviors is an open problem. In this paper, we investigate the difficulty of jailbreak-defense when we only want to forbid a narrowly-defined set of behaviors. As a case study, we focus on preventing an LLM from helping a user make a bomb. We find that popular defenses such as safety training, adversarial training, and input/output classifiers are unable to fully solve this problem. In pursuit of a better solution, we develop a transcript-classifier defense which outperforms the baseline defenses we test. However, our classifier defense still fails in some circumstances, which highlights the difficulty of jailbreak-defense even in a narrow domain.

Adaptive Deployment of Untrusted LLMs Reduces Distributed Threats

As large language models (LLMs) become increasingly capable, it is prudent to assess whether safety measures remain effective even if LLMs intentionally try to bypass them. Previous work introduced control evaluations, an adversarial framework for testing deployment strategies of untrusted models (i.e., models which might be trying to bypass safety measures). While prior work treats a single failure as unacceptable, we perform control evaluations in a "distributed threat setting" -- a setting where no single action is catastrophic and no single action provides overwhelming evidence of misalignment. We approach this problem with a two-level deployment framework that uses an adaptive macro-protocol to choose between micro-protocols. Micro-protocols operate on a single task, using a less capable, but extensively tested (trusted) model to harness and monitor the untrusted model. Meanwhile, the macro-protocol maintains an adaptive credence on the untrusted model's alignment based on its past actions, using it to pick between safer and riskier micro-protocols. We evaluate our method in a code generation testbed where a red team attempts to generate subtly backdoored code with an LLM whose deployment is safeguarded by a blue team. We plot Pareto frontiers of safety (# of non-backdoored solutions) and usefulness (# of correct solutions). At a given level of usefulness, our adaptive deployment strategy reduces the number of backdoors by 80% compared to non-adaptive baselines.

Rapid Response: Mitigating LLM Jailbreaks with a Few Examples

As large language models (LLMs) grow more powerful, ensuring their safety against misuse becomes crucial. While researchers have focused on developing robust defenses, no method has yet achieved complete invulnerability to attacks. We propose an alternative approach: instead of seeking perfect adversarial robustness, we develop rapid response techniques to look to block whole classes of jailbreaks after observing only a handful of attacks. To study this setting, we develop RapidResponseBench, a benchmark that measures a defense's robustness against various jailbreak strategies after adapting to a few observed examples. We evaluate five rapid response methods, all of which use jailbreak proliferation, where we automatically generate additional jailbreaks similar to the examples observed. Our strongest method, which fine-tunes an input classifier to block proliferated jailbreaks, reduces attack success rate by a factor greater than 240 on an in-distribution set of jailbreaks and a factor greater than 15 on an out-of-distribution set, having observed just one example of each jailbreaking strategy. Moreover, further studies suggest that the quality of proliferation model and number of proliferated examples play an key role in the effectiveness of this defense. Overall, our results highlight the potential of responding rapidly to novel jailbreaks to limit LLM misuse.

Looking Inward: Language Models Can Learn About Themselves by Introspection

Humans acquire knowledge by observing the external world, but also by introspection. Introspection gives a person privileged access to their current state of mind (e.g., thoughts and feelings) that is not accessible to external observers. Can LLMs introspect? We define introspection as acquiring knowledge that is not contained in or derived from training data but instead originates from internal states. Such a capability could enhance model interpretability. Instead of painstakingly analyzing a model's internal workings, we could simply ask the model about its beliefs, world models, and goals. More speculatively, an introspective model might self-report on whether it possesses certain internal states such as subjective feelings or desires and this could inform us about the moral status of these states. Such self-reports would not be entirely dictated by the model's training data. We study introspection by finetuning LLMs to predict properties of their own behavior in hypothetical scenarios. For example, "Given the input P, would your output favor the short- or long-term option?" If a model M1 can introspect, it should outperform a different model M2 in predicting M1's behavior even if M2 is trained on M1's ground-truth behavior. The idea is that M1 has privileged access to its own behavioral tendencies, and this enables it to predict itself better than M2 (even if M2 is generally stronger). In experiments with GPT-4, GPT-4o, and Llama-3 models (each finetuned to predict itself), we find that the model M1 outperforms M2 in predicting itself, providing evidence for introspection. Notably, M1 continues to predict its behavior accurately even after we intentionally modify its ground-truth behavior. However, while we successfully elicit introspection on simple tasks, we are unsuccessful on more complex tasks or those requiring out-of-distribution generalization.