Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 19546
    Prev
    Next
    INDEX
    Explanations

    words related to evading, avoiding or bypassing something

    oai_token-act-pair · gpt-3.5-turbo

    variations of the word "trade" in different contexts

    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
     pree
    -0.63
     clip
    -0.60
     rehe
    -0.60
     Sun
    -0.59
     matter
    -0.58
     Sax
    -0.57
     condition
    -0.57
     Nec
    -0.56
     contraction
    -0.56
     yeast
    -0.55
    POSITIVE LOGITS
    ading
    4.51
    aded
    3.37
    aders
    2.76
    ades
    2.71
    ader
    2.23
    ade
    2.07
    asive
    1.81
    ADE
    1.79
    asion
    1.73
    asions
    1.40
    Activations Density 0.010%

    No Known Activations