© 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. 4-RES-JB
    11. 2936
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to various political figures and entities

    oai_token-act-pair · gpt-3.5-turbo

    proper nouns related to nationality, political parties, and specific identities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    uador
    -0.84
    anwhile
    -0.72
    laus
    -0.70
    thora
    -0.69
    skirts
    -0.69
     Guan
    -0.69
    legraph
    -0.67
    ]'
    -0.66
    olitan
    -0.66
    andr
    -0.66
    POSITIVE LOGITS
     counterparts
    1.15
     counterpart
    1.00
     buddies
    0.85
     brethren
    0.84
     equivalents
    0.78
     cousins
    0.76
     career
    0.71
     cohorts
    0.70
     holdings
    0.69
     cousin
    0.69
    Activations Density 0.708%

    No Known Activations