© 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. 7677
    Prev
    Next
    INDEX
    Explanations

    mentions of a particular name or entity

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "Cap" or related variations, suggesting a focus on capitalization or specific entities represented by "Cap"

    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
    hower
    -0.95
    ĪĴ
    -0.86
    ¿½
    -0.84
     silence
    -0.79
    cause
    -0.74
    perse
    -0.70
    hood
    -0.69
     MSM
    -0.64
     anten
    -0.64
     td
    -0.63
    POSITIVE LOGITS
    itol
    1.43
    itals
    1.23
    acity
    1.19
    illary
    1.19
    rice
    1.17
    uchin
    1.09
    rices
    1.06
    abilities
    1.06
    itan
    1.00
    ulet
    0.99
    Activations Density 0.017%

    No Known Activations