© 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 768
    9. 8-RES_FS768-JB
    11. 754
    Prev
    Next
    INDEX
    Explanations

    the word "Any" coupled with various follow-up words or phrases

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    616
    +0.13
    0.6%
    484
    +0.12
    0.6%
    649
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    484
    +0.13
    0.09
    616
    +0.12
    0.08
    163
    +0.12
    0.07
    Negative Logits
    rex
    -0.76
    endas
    -0.73
    seless
    -0.70
    expensive
    -0.65
    rea
    -0.65
    itals
    -0.65
    went
    -0.61
    ually
    -0.60
    appings
    -0.59
    romy
    -0.59
    POSITIVE LOGITS
    THING
    1.55
     conceivable
    1.17
    WHERE
    1.15
    body
    1.12
    where
    1.07
     imaginable
    1.06
     semblance
    0.98
    thin
    0.94
    ONE
    0.93
     kind
    0.93
    Activations Density 1.824%

    No Known Activations