© 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. 1595
    Prev
    Next
    INDEX
    Explanations

    mentions of taking down or removing something

    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
    125
    +0.14
    0.6%
    262
    +0.10
    0.5%
    151
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    125
    +0.14
    0.04
    262
    +0.10
    0.03
    151
    +0.10
    0.03
    Negative Logits
    ugh
    -0.72
    kefeller
    -0.66
    gression
    -0.64
    xtap
    -0.62
    itia
    -0.62
     Russo
    -0.60
     Huck
    -0.60
    archives
    -0.58
     Kenya
    -0.56
    aido
    -0.56
    POSITIVE LOGITS
    graded
    1.21
    grading
    1.14
    grades
    1.03
    stairs
    0.97
    LOAD
    0.95
     stairs
    0.93
    hill
    0.83
    pour
    0.80
    size
    0.80
    sized
    0.79
    Activations Density 0.049%

    No Known Activations