© 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. 415
    Prev
    Next
    INDEX
    Explanations

    criminal activities and legal terms

    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
    153
    +0.10
    0.5%
    148
    +0.10
    0.5%
    593
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    199
    +0.10
    0.05
    711
    +0.10
    0.03
    628
    +0.09
    0.04
    Negative Logits
    soDeliveryDate
    -0.88
    DragonMagazine
    -0.82
    kefeller
    -0.79
    isSpecialOrderable
    -0.75
    NetMessage
    -0.74
     trillions
    -0.72
    mega
    -0.70
    Tok
    -0.69
     trickle
    -0.69
     trough
    -0.67
    POSITIVE LOGITS
     manslaughter
    1.31
     felony
    1.26
     aggravated
    1.17
     burglary
    1.15
     misdemeanor
    1.15
     felon
    1.10
     offences
    1.08
     perjury
    1.03
     unlawful
    1.02
     unlawfully
    1.00
    Activations Density 0.438%

    No Known Activations