© 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. 4-RES-JB
    11. 14915
    Prev
    Next
    INDEX
    Explanations

    instances of the word "avoid" along with the value 9 or 10, indicating an emphasis on caution or prevention

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "avoid."

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

    No Comments

    Negative Logits
    iop
    -0.90
    geist
    -0.69
    essee
    -0.68
    Rated
    -0.67
    cart
    -0.66
    ART
    -0.64
    otle
    -0.63
    opter
    -0.63
    Ready
    -0.62
     Directorate
    -0.62
    POSITIVE LOGITS
     detection
    0.78
    ably
    0.75
     pitfalls
    0.74
     wasting
    0.72
    ading
    0.68
     evade
    0.67
    avoid
    0.67
     avoidance
    0.66
    nels
    0.66
    azaki
    0.65
    Activations Density 0.026%

    No Known Activations