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. 3492
    Prev
    Next
    INDEX
    Explanations

    dates and locations expressed as cardinal numbers

    oai_token-act-pair · gpt-3.5-turbo

    negative connotations or mentions of conflict

    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
    advertisement
    -1.10
    ÂŃ
    -1.03
    Advertisements
    -0.84
    risome
    -0.77
    â̳
    -0.74
     â̦"
    -0.74
    â̦."
    -0.71
    â̲
    -0.69
    â̦â̦â̦â̦â̦â̦â̦â̦
    -0.69
    ADVERTISEMENT
    -0.69
    POSITIVE LOGITS
     -
    2.98
     --
    1.61
     �
    1.56
     ---
    1.41
    --
    1.35
     –
    1.25
     "-
    1.25
     (-
    1.22
     +
    1.21
     -->
    1.19
    Activations Density 0.061%

    No Known Activations