© 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 12288
    9. 8-RES_FS12288-JB
    11. 10323
    Prev
    Next
    INDEX
    Explanations

    names of famous individuals

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    530
    +0.14
    0.7%
    373
    +0.12
    0.6%
    326
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    530
    +0.14
    0.08
    326
    +0.12
    0.05
    373
    +0.11
    -0.02
    Negative Logits
    hex
    -0.72
    ciplinary
    -0.68
    ghai
    -0.66
    gate
    -0.64
    agate
    -0.64
    gebra
    -0.63
    icing
    -0.61
    tons
    -0.60
    coord
    -0.58
    apist
    -0.58
    POSITIVE LOGITS
    VILLE
    1.55
    TON
    1.44
    IAN
    1.42
    LEY
    1.39
    WOOD
    1.37
    SON
    1.34
    ION
    1.33
    URA
    1.33
    ANS
    1.32
    INE
    1.32
    Activations Density 0.211%

    No Known Activations