Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 1536
    5. 8-RES_FS1536-JB
    6. 1370
    Prev
    Next
    INDEX
    Explanations

    information about crimes and law enforcement 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_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    31
    +0.12
    0.5%
    508
    +0.10
    0.5%
    537
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    31
    +0.12
    0.11
    752
    +0.10
    0.11
    508
    +0.09
    0.09
    Negative Logits
    furt
    -0.87
    acles
    -0.86
    aeus
    -0.83
    aphael
    -0.76
    isks
    -0.75
    ihad
    -0.74
    Sov
    -0.74
    esp
    -0.73
    enstein
    -0.73
    iable
    -0.72
    POSITIVE LOGITS
     Sheriff
    1.52
     sheriff
    1.43
     County
    1.22
     commissioners
    1.19
     coroner
    1.18
     Supervisor
    1.18
     Clerk
    1.17
     Commissioners
    1.15
     deputies
    1.15
     commissioner
    1.12
    Activations Density 2.215%

    No Known Activations