© 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. 0-RES-JB
    11. 574
    Prev
    Next
    INDEX
    Explanations

    statements of declining to provide information or comment

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "declined" or similar terms related to refusal or rejection

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     sym
    -0.69
     stirring
    -0.62
     alignment
    -0.61
     Mak
    -0.57
     Teacher
    -0.56
     mash
    -0.56
     clone
    -0.56
     knowing
    -0.56
     glue
    -0.56
     imagin
    -0.56
    POSITIVE LOGITS
     declined
    3.61
     declines
    2.18
     refused
    1.81
     declining
    1.77
     decline
    1.70
     plummeted
    1.48
     denied
    1.31
     dropped
    1.30
     ceased
    1.29
     retreated
    1.29
    Activations Density 0.012%

    No Known Activations