© 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. 10316
    Prev
    Next
    INDEX
    Explanations

    technical terms related to software development processes and concepts

    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
    93
    +0.10
    0.5%
    503
    +0.10
    0.5%
    299
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.06
    763
    +0.10
    0.04
    503
    +0.10
    0.03
    Negative Logits
    waves
    -0.70
    BOOK
    -0.68
     Belt
    -0.67
    foot
    -0.66
    GAME
    -0.64
     LORD
    -0.64
    BALL
    -0.64
    water
    -0.64
    dress
    -0.63
    shock
    -0.63
    POSITIVE LOGITS
    ating
    2.17
    ated
    2.01
    ators
    1.99
    ations
    1.97
    ator
    1.86
    ates
    1.84
    ational
    1.81
    atory
    1.77
    ative
    1.77
    ate
    1.74
    Activations Density 0.228%

    No Known Activations