© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 1191
    Prev
    Next
    INDEX
    Explanations

    technical terms and specialized jargon in academic or research texts

    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
    149
    +0.10
    0.5%
    288
    +0.10
    0.4%
    503
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.10
    0.04
    232
    +0.10
    0.04
    503
    +0.10
    0.03
    Negative Logits
     Spice
    -0.65
     miscarriage
    -0.65
     away
    -0.64
     turkey
    -0.64
     initiation
    -0.64
    ħĭ
    -0.64
     inclination
    -0.64
    PRESS
    -0.63
     slee
    -0.62
     spin
    -0.62
    POSITIVE LOGITS
    ified
    2.11
    ifying
    2.05
    ification
    1.92
    ifier
    1.86
    ifies
    1.83
    ifiers
    1.83
    ifications
    1.76
    ifiable
    1.71
    ities
    1.60
    itatively
    1.54
    Activations Density 0.113%

    No Known Activations