© 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. 1591
    Prev
    Next
    INDEX
    Explanations

    words related to food texture, specifically focusing on something being crispy or having a crispy texture

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    361
    +0.12
    0.5%
    715
    +0.11
    0.5%
    526
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.12
    0.01
    555
    +0.11
    0.01
    288
    +0.10
    0.01
    Negative Logits
    gging
    -0.67
    isSpecialOrderable
    -0.63
    WAY
    -0.63
    Footnote
    -0.62
    JV
    -0.62
    YL
    -0.62
    ACT
    -0.61
    WAYS
    -0.60
     Graveyard
    -0.59
    adder
    -0.59
    POSITIVE LOGITS
    iness
    0.94
    bread
    0.92
    essen
    0.88
    ulate
    0.88
    arity
    0.85
    ulent
    0.82
    acies
    0.81
     crisp
    0.81
    ness
    0.76
    batch
    0.74
    Activations Density 0.033%

    No Known Activations