Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 46075
    Prev
    Next
    INDEX
    Explanations

    references to political figures and events

    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
    480
    +0.46
    2.4%
    481
    +0.14
    0.7%
    420
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.46
    0.07
    346
    +0.14
    0.06
    146
    +0.12
    0.06
    Negative Logits
    âĢ¢âĢ¢âĢ¢âĢ¢
    -0.68
    $.
    -0.67
    ãĥł
    -0.63
    âĶĢâĶĢâĶĢâĶĢ
    -0.62
    selves
    -0.59
    Translation
    -0.59
    Occ
    -0.58
    Eva
    -0.56
    Els
    -0.56
    ————————————————
    -0.55
    POSITIVE LOGITS
     reacted
    1.09
     meanwhile
    1.08
     intervened
    0.95
     spokesman
    0.92
     congratulated
    0.92
     applauded
    0.92
     responded
    0.90
     welcomed
    0.89
     spokeswoman
    0.88
     mourn
    0.85
    Activations Density 0.479%

    No Known Activations