© 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 49152
    9. 8-RES_FS49152-JB
    11. 37236
    Prev
    Next
    INDEX
    Explanations

    numerical ratings or scores often related to performance or assessment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    71
    +0.13
    0.6%
    127
    +0.12
    0.5%
    385
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.13
    0.02
    310
    +0.12
    0.02
    488
    +0.10
    0.02
    Negative Logits
    uden
    -0.83
    hood
    -0.74
    matically
    -0.70
    lie
    -0.66
    vati
    -0.66
    gets
    -0.65
    tons
    -0.65
    mes
    -0.64
    arted
    -0.64
    teen
    -0.63
    POSITIVE LOGITS
    ASC
    0.85
    çĶ°
    0.82
    80
    0.80
     cents
    0.79
    REM
    0.78
    IRC
    0.76
    01
    0.74
    70
    0.73
    803
    0.72
    CTV
    0.72
    Activations Density 0.027%

    No Known Activations