© 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. 23907
    Prev
    Next
    INDEX
    Explanations

    instances of the word "respond" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered 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
    195
    +0.13
    0.6%
    509
    +0.10
    0.4%
    759
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    195
    +0.13
    0.02
    70
    +0.10
    0.03
    624
    +0.10
    0.03
    Negative Logits
    dar
    -0.70
    colo
    -0.65
    fre
    -0.65
    thing
    -0.62
     Marin
    -0.62
     Demons
    -0.62
     marrow
    -0.60
    fi
    -0.60
     Thieves
    -0.60
     groove
    -0.57
    POSITIVE LOGITS
     favorably
    1.12
     thereto
    1.07
     positively
    1.07
    ivated
    1.02
     negatively
    0.91
     angrily
    0.90
     appropriately
    0.89
     harshly
    0.86
     adequately
    0.86
    ively
    0.85
    Activations Density 0.030%

    No Known Activations