© 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. 37333
    Prev
    Next
    INDEX
    Explanations

    negative implications or consequences associated with various situations or actions

    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
    102
    +0.16
    0.8%
    295
    +0.11
    0.5%
    747
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    102
    +0.16
    0.06
    152
    +0.11
    0.05
    124
    +0.11
    0.05
    Negative Logits
    ulhu
    -0.76
     Flavoring
    -0.72
    iage
    -0.71
     Transparency
    -0.68
     deduction
    -0.67
     clause
    -0.67
     SHARES
    -0.66
     Curve
    -0.65
     tweet
    -0.64
    vernment
    -0.64
    POSITIVE LOGITS
    compatible
    1.18
    enough
    1.11
    eligible
    1.08
    aware
    1.02
    focused
    1.01
    tested
    1.01
    producing
    0.98
    eyed
    0.98
    dependent
    0.94
    years
    0.94
    Activations Density 0.074%

    No Known Activations