© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 49152
    9. 8-RES_FS49152-JB
    11. 41733
    Prev
    Next
    INDEX
    Explanations

    references to political leaders and government officials

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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

    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.20
    1.1%
    87
    +0.19
    1.0%
    407
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.20
    0.05
    136
    +0.19
    0.05
    538
    +0.09
    0.04
    Negative Logits
    bleacher
    -0.62
    EStreamFrame
    -0.60
    amily
    -0.50
    xual
    -0.50
    zzle
    -0.49
    Äĩ
    -0.49
    zsche
    -0.47
    knit
    -0.47
     COUNTY
    -0.47
    orsi
    -0.47
    POSITIVE LOGITS
     spokesman
    0.56
     congratulated
    0.53
    ofi
    0.53
     Lore
    0.51
     reacted
    0.50
     Rupert
    0.50
     reiterated
    0.50
     meanwhile
    0.49
     Vincent
    0.49
    assies
    0.49
    Activations Density 0.227%

    No Known Activations