© 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. 6798
    Prev
    Next
    INDEX
    Explanations

    action words related to technology and coding

    oai_token-act-pair · gpt-3.5-turbo

    references to observation or attention in a descriptive context

    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
     Reincarnated
    -0.75
    netflix
    -0.72
    ablishment
    -0.69
    oppable
    -0.64
    SAN
    -0.63
    NetMessage
    -0.63
    SPONSORED
    -0.62
    existence
    -0.61
    Ħ¢
    -0.61
    à¨
    -0.61
    POSITIVE LOGITS
     however
    1.20
     note
    0.95
     caveat
    0.91
     also
    0.88
     caveats
    0.87
     additionally
    0.86
     therefore
    0.83
     furthermore
    0.82
     NOTE
    0.82
     tho
    0.82
    Activations Density 0.757%

    No Known Activations