© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 18516
    Prev
    Next
    INDEX
    Explanations

    terms related to negative characteristics or behaviors, particularly related to ignorance, hypocrisy, and dishonesty

    oai_token-act-pair · gpt-3.5-turbo

    suffixes that indicate various forms of criticism or negative traits

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    nesium
    -0.80
    ENN
    -0.73
    ITNESS
    -0.69
    ERC
    -0.69
     mosqu
    -0.63
    onga
    -0.63
     cryst
    -0.62
    ergy
    -0.62
    yton
    -0.61
    jam
    -0.61
    POSITIVE LOGITS
    ously
    1.15
    iness
    0.97
    ous
    0.95
     ridden
    0.92
    uous
    0.89
     plag
    0.86
     perpetrated
    0.86
    ality
    0.86
    ulence
    0.86
     quot
    0.86
    Activations Density 0.206%

    No Known Activations