© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 49152
    9. 8-RES_FS49152-JB
    11. 37158
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "clip" in various contexts, indicating a focus on video or audio clips

    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
    187
    +0.13
    0.6%
    276
    +0.10
    0.5%
    241
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    187
    +0.13
    0.01
    276
    +0.10
    0.01
    537
    +0.10
    0.01
    Negative Logits
    ichick
    -0.94
    iott
    -0.94
    ccording
    -0.74
    ADRA
    -0.69
     Schwarz
    -0.66
    velt
    -0.66
    ksh
    -0.65
    orney
    -0.62
    hips
    -0.61
    estate
    -0.61
    POSITIVE LOGITS
    clips
    1.08
     clip
    1.05
     clips
    1.05
    clip
    1.01
    boards
    1.00
     Clip
    0.90
     footage
    0.88
    scenes
    0.84
    ography
    0.83
    ograp
    0.81
    Activations Density 0.006%

    No Known Activations