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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 41625
    Prev
    Next
    INDEX
    Explanations

    mathematical symbols and expressions

    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
    640
    +0.19
    0.9%
    160
    +0.15
    0.7%
    373
    +0.15
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.19
    0.05
    160
    +0.15
    0.04
    6
    +0.15
    0.04
    Negative Logits
    hement
    -0.89
    rity
    -0.75
     lifes
    -0.71
     spearheaded
    -0.71
     Borderlands
    -0.69
     exha
    -0.69
     Volunteers
    -0.68
     giveaways
    -0.68
     stren
    -0.68
     relocation
    -0.67
    POSITIVE LOGITS
    $$
    0.97
    pmwiki
    0.97
    "â̦
    0.94
    shall
    0.93
    Quote
    0.91
    ³³³³³³³³³³³³³³³³
    0.89
    _(
    0.85
    λ
    0.85
    #$
    0.84
    Î
    0.83
    Activations Density 0.071%

    No Known Activations