© 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. 183
    Prev
    Next
    INDEX
    Explanations

    quotation marks with specific associated words or phrases inside them

    oai_token-act-pair · gpt-3.5-turbo

    quotation marks and phrases enclosed within them

    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
     Vog
    -0.74
     Archdemon
    -0.74
     Concord
    -0.73
     Nieto
    -0.73
     cardinal
    -0.72
     affiliate
    -0.72
     chair
    -0.71
     Tribune
    -0.71
     succession
    -0.71
     candles
    -0.70
    POSITIVE LOGITS
    normal
    1.45
    official
    1.37
    cheat
    1.33
    natural
    1.33
    safe
    1.32
    feel
    1.32
    pure
    1.30
    true
    1.30
    traditional
    1.30
    real
    1.28
    Activations Density 0.122%

    No Known Activations