© 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 12288
    9. 8-RES_FS12288-JB
    11. 1331
    Prev
    Next
    INDEX
    Explanations

    the word "null" used in a context related to voiding or invalidating 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_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    739
    +0.14
    0.6%
    158
    +0.10
    0.5%
    503
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.14
    0.04
    326
    +0.10
    0.03
    232
    +0.09
    0.03
    Negative Logits
    ieri
    -0.78
     Harlem
    -0.73
    nings
    -0.72
    ufact
    -0.70
     Lans
    -0.68
    ches
    -0.67
    ducers
    -0.64
    naissance
    -0.64
    ahon
    -0.63
    oses
    -0.61
    POSITIVE LOGITS
    ifying
    0.97
    ified
    0.94
    ptr
    0.94
    ifies
    0.90
    ification
    0.85
    erd
    0.82
    ify
    0.82
    ifier
    0.82
    ensible
    0.81
    mate
    0.81
    Activations Density 0.120%

    No Known Activations