© 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. 528
    Prev
    Next
    INDEX
    Explanations

    words related to specific characters or entities, especially names

    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
    494
    +0.12
    0.5%
    436
    +0.11
    0.5%
    553
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.12
    0.06
    232
    +0.11
    0.06
    299
    +0.10
    0.05
    Negative Logits
    atures
    -0.74
    ensical
    -0.67
    sylv
    -0.67
    natureconservancy
    -0.65
     timeless
    -0.62
    bral
    -0.62
    ersen
    -0.61
    akeru
    -0.59
    wives
    -0.58
    SPONSORED
    -0.58
    POSITIVE LOGITS
    ooth
    0.85
    adesh
    0.79
    istan
    0.74
     Reb
    0.74
    vous
    0.74
    ر
    0.70
    obin
    0.69
    bash
    0.69
     Khalid
    0.66
    Ö¼
    0.66
    Activations Density 0.418%

    No Known Activations