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. 26313
    Prev
    Next
    INDEX
    Explanations

    terms related to sports games and political elections

    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
    281
    +0.10
    0.5%
    310
    +0.10
    0.5%
    383
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.10
    0.05
    749
    +0.10
    0.03
    281
    +0.10
    0.03
    Negative Logits
     collaborators
    -0.72
    anguages
    -0.70
    recy
    -0.68
    ocate
    -0.65
    unia
    -0.65
     journals
    -0.64
     publications
    -0.64
     collaborating
    -0.63
    laws
    -0.63
    coon
    -0.62
    POSITIVE LOGITS
     postponed
    1.07
     replay
    0.93
     forfeit
    0.91
     televised
    0.89
     Finals
    0.84
     undecided
    0.82
     scheduled
    0.81
     theat
    0.79
     adjourn
    0.76
     spectators
    0.75
    Activations Density 0.232%

    No Known Activations