© 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 768
    9. 8-RES_FS768-JB
    11. 726
    Prev
    Next
    INDEX
    Explanations

    stories or articles about crimes and criminal activities

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    373
    +0.11
    0.5%
    363
    +0.10
    0.5%
    672
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    610
    +0.11
    0.09
    373
    +0.10
    -0.08
    363
    +0.10
    0.09
    Negative Logits
     Scrib
    -0.77
     Mechdragon
    -0.75
    Offline
    -0.61
     Tarant
    -0.61
     Thro
    -0.60
    TEXTURE
    -0.59
     NP
    -0.59
     vo
    -0.58
     Tac
    -0.56
     Noise
    -0.56
    POSITIVE LOGITS
    olds
    0.89
    aunts
    0.87
    old
    0.85
    appointed
    0.77
    umn
    0.73
    OLD
    0.72
    ago
    0.71
    young
    0.69
    older
    0.69
    gins
    0.68
    Activations Density 4.701%

    No Known Activations