© 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. 9406
    Prev
    Next
    INDEX
    Explanations

    Portuguese language-specific phrases and names

    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
    690
    +0.13
    0.6%
    610
    +0.12
    0.6%
    674
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    674
    +0.13
    0.03
    93
    +0.12
    0.03
    545
    +0.11
    0.02
    Negative Logits
    ulence
    -0.74
     behavi
    -0.71
     ethics
    -0.69
     tremend
    -0.68
    NESS
    -0.68
    lihood
    -0.66
     neigh
    -0.64
     prosper
    -0.64
    nos
    -0.63
    sense
    -0.62
    POSITIVE LOGITS
     Played
    1.19
    pleted
    1.10
     Bought
    1.07
    Used
    1.04
    ired
    1.02
    ceived
    1.02
    Posted
    1.01
    tained
    1.01
    wrote
    1.00
     Started
    0.99
    Activations Density 0.190%

    No Known Activations