© 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. 20003
    Prev
    Next
    INDEX
    Explanations

    dialogue excerpts and commentary

    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
    393
    +0.12
    0.6%
    685
    +0.10
    0.5%
    297
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    297
    +0.12
    0.04
    43
    +0.10
    0.04
    389
    +0.10
    0.04
    Negative Logits
    STDOUT
    -0.72
    adas
    -0.71
    ady
    -0.66
    ulz
    -0.63
    Associated
    -0.63
    SourceFile
    -0.63
    PDF
    -0.63
    AFP
    -0.62
    osite
    -0.61
    interstitial
    -0.61
    POSITIVE LOGITS
     yeah
    1.41
     uh
    1.05
     I
    1.04
     yes
    1.00
     sir
    0.99
     hey
    0.96
     please
    0.92
     secondly
    0.90
     wow
    0.90
     um
    0.89
    Activations Density 0.182%

    No Known Activations