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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 43390
    Prev
    Next
    INDEX
    Explanations

    references to fans of various sports teams and games

    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
    39
    +0.13
    0.6%
    502
    +0.11
    0.5%
    242
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    568
    +0.13
    0.03
    91
    +0.11
    0.03
    575
    +0.11
    0.03
    Negative Logits
     divisions
    -0.68
    ãģĤ
    -0.62
     embassies
    -0.61
     penalties
    -0.61
     kingdoms
    -0.60
    men
    -0.60
    runners
    -0.60
    MEN
    -0.60
     balls
    -0.60
     dancers
    -0.59
    POSITIVE LOGITS
     extraord
    0.96
    digy
    0.91
     myself
    0.81
    rette
    0.77
    turned
    0.76
    abilia
    0.76
     rejoice
    0.75
    chuk
    0.73
     nonetheless
    0.71
     blogger
    0.70
    Activations Density 0.065%

    No Known Activations