© 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 49152
    9. 8-RES_FS49152-JB
    11. 37197
    Prev
    Next
    INDEX
    Explanations

    expressions of uncertainty or subjectivity about situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    373
    +0.14
    0.7%
    195
    +0.10
    0.5%
    544
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    195
    +0.14
    0.03
    544
    +0.10
    0.03
    183
    +0.10
    0.03
    Negative Logits
    ests
    -0.82
    rouse
    -0.78
    orem
    -0.75
    otos
    -0.75
    andise
    -0.74
    ilts
    -0.71
    pez
    -0.69
    izons
    -0.68
    perature
    -0.68
    venge
    -0.68
    POSITIVE LOGITS
     unlikely
    0.93
     doubtful
    0.91
     probable
    0.78
     unclear
    0.75
     prudent
    0.72
     plausible
    0.71
     advisable
    0.70
     folly
    0.70
     feasible
    0.70
     imperative
    0.69
    Activations Density 0.035%

    No Known Activations