© 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. 37269
    Prev
    Next
    INDEX
    Explanations

    action words related to obtaining or collecting resources

    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
    542
    +0.11
    0.5%
    145
    +0.11
    0.5%
    383
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    217
    +0.11
    0.03
    542
    +0.11
    0.03
    677
    +0.10
    0.03
    Negative Logits
    atever
    -0.76
    rium
    -0.65
    mare
    -0.64
    repeat
    -0.59
    leader
    -0.58
    BLE
    -0.57
    Demand
    -0.57
    Rate
    -0.56
    meter
    -0.56
    olas
    -0.56
    POSITIVE LOGITS
     from
    0.91
     anonymously
    0.87
     incidentally
    0.82
     by
    0.82
     cheaply
    0.81
     BY
    0.77
     via
    0.76
     aback
    0.76
     illegally
    0.75
     FROM
    0.73
    Activations Density 0.058%

    No Known Activations