© 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 12288
    9. 8-RES_FS12288-JB
    11. 469
    Prev
    Next
    INDEX
    Explanations

    strong adjectives and phrases related to identification or perception

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    408
    +0.18
    0.8%
    462
    +0.10
    0.5%
    284
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    408
    +0.18
    0.06
    284
    +0.10
    0.03
    763
    +0.09
    0.03
    Negative Logits
    cius
    -0.75
    alian
    -0.73
    dm
    -0.70
     Originally
    -0.70
    sometimes
    -0.70
    prone
    -0.70
     Beware
    -0.69
    }{
    -0.67
    rax
    -0.67
    ometimes
    -0.67
    POSITIVE LOGITS
     explanation
    1.23
     indication
    1.19
     reason
    1.11
     recourse
    1.06
     whatsoever
    1.05
     distinction
    0.99
     excuse
    0.99
     reperc
    0.94
     motive
    0.93
     justification
    0.93
    Activations Density 0.211%

    No Known Activations