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. 41204
    Prev
    Next
    INDEX
    Explanations

    references to UFC events and organizations

    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
    288
    +0.12
    0.5%
    496
    +0.11
    0.5%
    146
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    290
    +0.12
    0.02
    146
    +0.11
    0.02
    754
    +0.10
    0.01
    Negative Logits
    matter
    -0.82
    Reviewer
    -0.75
    anyahu
    -0.73
     Pelosi
    -0.72
     MacArthur
    -0.71
    abee
    -0.68
    itia
    -0.68
    combe
    -0.68
    mary
    -0.68
    zzle
    -0.68
    POSITIVE LOGITS
     Tournament
    0.88
     Fight
    0.84
     Tonight
    0.81
    Fighting
    0.80
     Champion
    0.80
     heavyweight
    0.79
     tournament
    0.79
     Championships
    0.78
     arena
    0.76
     Network
    0.74
    Activations Density 0.007%

    No Known Activations