© 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. 417
    Prev
    Next
    INDEX
    Explanations

    slashes followed by numbers reminiscent of a date and a medical term or event of significance

    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
    641
    +0.15
    0.7%
    373
    +0.10
    0.5%
    291
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    350
    +0.15
    0.04
    641
    +0.10
    0.04
    127
    +0.09
    0.03
    Negative Logits
    abase
    -0.63
     Builder
    -0.63
    ause
    -0.62
    ħĭ
    -0.61
     Masquerade
    -0.60
     Conversation
    -0.59
     tert
    -0.59
     Appl
    -0.58
     Compass
    -0.58
     Diss
    -0.57
    POSITIVE LOGITS
    999
    0.84
    11
    0.79
    Downloadha
    0.75
    woods
    0.75
    576
    0.73
    umerable
    0.72
    996
    0.70
    monary
    0.70
    9999
    0.70
    warts
    0.68
    Activations Density 0.183%

    No Known Activations