© 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. 37153
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "today."

    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
    254
    +0.10
    0.5%
    709
    +0.10
    0.5%
    688
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    254
    +0.10
    0.02
    688
    +0.10
    0.02
    651
    +0.10
    0.02
    Negative Logits
    ashtra
    -0.70
    aughed
    -0.69
    rosse
    -0.66
     Hammond
    -0.61
     Conquer
    -0.61
     tyr
    -0.58
     Leth
    -0.58
    emis
    -0.57
    Eth
    -0.57
    este
    -0.55
    POSITIVE LOGITS
    days
    0.86
    's
    0.85
    utical
    0.76
    care
    0.75
    âĢ²
    0.73
    astical
    0.69
    abouts
    0.69
    dream
    0.66
    stall
    0.66
    break
    0.65
    Activations Density 0.028%

    No Known Activations