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 768
    5. 8-RES_FS768-JB
    6. 335
    Prev
    Next
    INDEX
    Explanations

    names of individuals involved in legal disputes or conflict situations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    530
    +0.15
    0.7%
    326
    +0.14
    0.7%
    373
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    530
    +0.15
    0.13
    326
    +0.14
    0.10
    373
    +0.11
    -0.05
    Negative Logits
    hex
    -0.65
    ks
    -0.63
    perties
    -0.63
    unia
    -0.62
    lahoma
    -0.62
    gy
    -0.62
    idae
    -0.61
    ge
    -0.60
    opol
    -0.60
    acity
    -0.59
    POSITIVE LOGITS
    ING
    1.73
    ION
    1.61
    WOR
    1.59
    ER
    1.58
    HEAD
    1.58
    MENT
    1.58
    ERS
    1.58
    LY
    1.57
    NESS
    1.57
    ABLE
    1.56
    Activations Density 1.370%

    No Known Activations