© 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 24576
    9. 8-RES_FS24576-JB
    11. 19834
    Prev
    Next
    INDEX
    Explanations

    words related to physical actions or objects related to a hit or a blow

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    526
    +0.12
    0.6%
    596
    +0.12
    0.5%
    62
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.01
    526
    +0.12
    0.01
    181
    +0.11
    0.01
    Negative Logits
    xual
    -0.74
    trl
    -0.71
    restricted
    -0.66
    nces
    -0.66
    士
    -0.64
     Journals
    -0.63
    es
    -0.63
     oath
    -0.62
    Parameters
    -0.60
    agents
    -0.60
    POSITIVE LOGITS
    butt
    1.11
     cheeks
    1.02
    plug
    0.98
    ocks
    0.93
    strap
    0.91
     cheek
    0.85
     butt
    0.84
    ock
    0.84
    iary
    0.83
    odox
    0.80
    Activations Density 0.025%

    No Known Activations