© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 10116
    Prev
    Next
    INDEX
    Explanations

    mentions of actions where someone did not comment or provide a response

    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
    148
    +0.10
    0.5%
    116
    +0.10
    0.5%
    698
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    455
    +0.10
    0.05
    645
    +0.10
    0.04
    434
    +0.09
    0.04
    Negative Logits
     doubtless
    -0.71
     Tags
    -0.69
    ngth
    -0.66
    akeru
    -0.66
    livion
    -0.65
    oir
    -0.64
     gorgeous
    -0.63
     wond
    -0.63
     beautifully
    -0.63
    eways
    -0.62
    POSITIVE LOGITS
     condone
    0.98
     prejud
    0.84
    icably
    0.82
     necessarily
    0.82
     regrets
    0.80
     comment
    0.80
     regret
    0.79
     personally
    0.78
     endorse
    0.77
     disrespect
    0.75
    Activations Density 0.240%

    No Known Activations