© 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. 9498
    Prev
    Next
    INDEX
    Explanations

    past tense verbs that denote action or decision-making

    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
    480
    +0.24
    1.2%
    447
    +0.23
    1.1%
    481
    +0.20
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    42
    +0.24
    0.04
    464
    +0.23
    0.04
    85
    +0.20
    0.04
    Negative Logits
     Niet
    -0.70
     Seym
    -0.69
     Azerb
    -0.66
    stood
    -0.64
     prest
    -0.63
    isphere
    -0.63
    edIn
    -0.61
    Reason
    -0.57
     Ingram
    -0.56
    istani
    -0.56
    POSITIVE LOGITS
    metics
    0.71
    Ĥª
    0.69
    actionDate
    0.69
    ync
    0.63
    zbollah
    0.62
    atem
    0.60
    malink
    0.58
    §
    0.57
     ][
    0.57
    ittens
    0.57
    Activations Density 0.122%

    No Known Activations