© 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. 1604
    Prev
    Next
    INDEX
    Explanations

    examples that are open-ended and related to case briefs or legal scenarios

    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
    214
    +0.10
    0.4%
    348
    +0.09
    0.4%
    540
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    348
    +0.10
    0.03
    540
    +0.09
    0.03
    370
    +0.09
    0.02
    Negative Logits
     livest
    -0.71
    nesday
    -0.71
     destro
    -0.68
    estern
    -0.68
     holiest
    -0.66
    onder
    -0.66
     phased
    -0.65
    hops
    -0.65
    orks
    -0.64
    rush
    -0.63
    POSITIVE LOGITS
    esan
    0.87
    cases
    0.82
     Keen
    0.81
    cmp
    0.81
    backs
    0.76
    suit
    0.70
    xia
    0.70
    ouf
    0.69
    anova
    0.69
    arella
    0.69
    Activations Density 0.038%

    No Known Activations