© 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 49152
    9. 8-RES_FS49152-JB
    11. 37265
    Prev
    Next
    INDEX
    Explanations

    words related to professions or roles in various fields

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    481
    +0.13
    0.6%
    679
    +0.11
    0.5%
    157
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    330
    +0.13
    0.04
    32
    +0.11
    0.04
    224
    +0.11
    0.04
    Negative Logits
    SpaceEngineers
    -0.72
    ãĥª
    -0.72
    Accessory
    -0.71
    ASED
    -0.68
    Ore
    -0.67
    SPONSORED
    -0.67
    UCT
    -0.67
    shapeshifter
    -0.62
    THIS
    -0.61
    ãĥĥãĥī
    -0.61
    POSITIVE LOGITS
    hip
    1.02
    hips
    0.93
     beware
    0.91
    cript
    0.90
    wana
    0.86
     Picks
    0.85
    cius
    0.84
     reacted
    0.81
    kaya
    0.81
    afety
    0.80
    Activations Density 0.091%

    No Known Activations