© 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. 882
    Prev
    Next
    INDEX
    Explanations

    facts or information stated as truth

    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
    213
    +0.12
    0.5%
    688
    +0.10
    0.5%
    651
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    213
    +0.12
    0.03
    78
    +0.10
    0.03
    236
    +0.10
    0.03
    Negative Logits
    avorite
    -0.68
     streng
    -0.67
    wana
    -0.64
    idden
    -0.62
    ESE
    -0.61
    itsch
    -0.61
    sung
    -0.60
    asca
    -0.60
    airs
    -0.60
    ascus
    -0.59
    POSITIVE LOGITS
    uality
    1.14
    ually
    1.12
    ional
    1.10
    orial
    1.06
    oids
    0.83
    finder
    0.81
    ual
    0.80
    fulness
    0.80
    itious
    0.79
    uation
    0.77
    Activations Density 0.056%

    No Known Activations