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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 49152
    5. 8-RES_FS49152-JB
    6. 25812
    Prev
    Next
    INDEX
    Explanations

    past tense verbs indicating actions or states

    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
    297
    +0.13
    0.6%
    566
    +0.10
    0.5%
    418
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    102
    +0.13
    0.09
    55
    +0.10
    0.09
    566
    +0.10
    0.08
    Negative Logits
     Correction
    -0.77
     Compare
    -0.73
     IMAGES
    -0.72
     Explain
    -0.70
     Ago
    -0.68
     Previous
    -0.68
     RELEASE
    -0.66
     Visit
    -0.65
    Update
    -0.65
    inas
    -0.65
    POSITIVE LOGITS
     destined
    1.00
     born
    0.96
     poised
    0.92
     conceived
    0.90
     able
    0.89
     indeed
    0.88
    wolves
    0.88
    nt
    0.87
    wolf
    0.83
     decidedly
    0.83
    Activations Density 0.282%

    No Known Activations