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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 25673
    Prev
    Next
    INDEX
    Explanations

    significant and noteworthy ideas or topics

    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
    53
    +0.12
    0.5%
    737
    +0.11
    0.5%
    264
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    737
    +0.12
    0.03
    588
    +0.11
    0.02
    531
    +0.11
    0.02
    Negative Logits
    hement
    -0.69
    lations
    -0.67
    chairs
    -0.65
    oths
    -0.64
    mbuds
    -0.63
    asers
    -0.62
    onds
    -0.61
     downed
    -0.61
    ushes
    -0.61
    urses
    -0.61
    POSITIVE LOGITS
    piring
    0.81
    ?,
    0.72
    ?:
    0.71
     happening
    0.70
    ?",
    0.68
     Enough
    0.67
    SPONSORED
    0.66
    est
    0.65
    iating
    0.64
    EST
    0.63
    Activations Density 0.070%

    No Known Activations