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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 768
    5. 8-RES_FS768-JB
    6. 623
    Prev
    Next
    INDEX
    Explanations

    legislation and political terms and phrases

    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
    591
    +0.10
    0.4%
    746
    +0.10
    0.4%
    555
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.10
    0.16
    288
    +0.10
    0.19
    555
    +0.09
    0.13
    Negative Logits
     Origin
    -0.74
     Dome
    -0.63
     Butt
    -0.60
     Io
    -0.60
     Penet
    -0.60
     Panther
    -0.60
     LCS
    -0.59
     Shack
    -0.59
     bearer
    -0.58
     HIP
    -0.58
    POSITIVE LOGITS
    acies
    1.26
    eties
    1.20
    ating
    1.15
    ements
    1.14
    ancies
    1.14
    ctions
    1.14
    ations
    1.13
    acy
    1.13
    atory
    1.13
    iencies
    1.12
    Activations Density 3.657%

    No Known Activations