© 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. 3157
    Prev
    Next
    INDEX
    Explanations

    political affiliations and characteristics such as race, religion, and legality

    oai_token-act-pair · gpt-3.5-turbo

    politically charged terms related to party affiliations and demographics

    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
    ĺħ
    -0.77
    stals
    -0.74
    redits
    -0.70
    confidence
    -0.70
    omics
    -0.70
    igers
    -0.70
    ainer
    -0.69
    runners
    -0.68
    earchers
    -0.68
    center
    -0.67
    POSITIVE LOGITS
     sounding
    1.00
    isable
    0.94
    izable
    0.83
    istic
    0.78
    ifiable
    0.76
    idious
    0.73
     nature
    0.73
     manner
    0.73
     crowds
    0.71
     offspring
    0.69
    Activations Density 0.303%

    No Known Activations