© 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 24576
    9. 8-RES_FS24576-JB
    11. 18778
    Prev
    Next
    INDEX
    Explanations

    past-tense action verbs

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    48
    +0.11
    0.5%
    2
    +0.10
    0.5%
    172
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    604
    +0.11
    0.05
    566
    +0.10
    0.06
    570
    +0.10
    0.02
    Negative Logits
    currently
    -0.75
    described
    -0.66
     presently
    -0.64
    enei
    -0.63
    mania
    -0.63
    Bah
    -0.62
    arel
    -0.61
     currently
    -0.61
     continued
    -0.60
    linked
    -0.59
    POSITIVE LOGITS
     beforehand
    0.76
    ãĤ¤ãĥĪ
    0.76
     LAST
    0.74
     yesterday
    0.70
     Doct
    0.67
     wrong
    0.67
     BEFORE
    0.67
    nesday
    0.66
     last
    0.66
    wolf
    0.62
    Activations Density 0.361%

    No Known Activations