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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    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. 25881
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate intention or future plans

    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
    373
    +0.17
    0.8%
    294
    +0.09
    0.4%
    242
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.17
    0.01
    373
    +0.09
    -0.00
    484
    +0.09
    0.01
    Negative Logits
    ICLE
    -0.65
    CVE
    -0.62
    insured
    -0.61
     persisted
    -0.60
    ¶
    -0.56
     mand
    -0.55
    mods
    -0.55
    paralle
    -0.54
    enance
    -0.54
    Glass
    -0.54
    POSITIVE LOGITS
     explore
    0.69
    ousy
    0.64
    BuyableInstoreAndOnline
    0.64
    \\\\\\\\
    0.64
     Siberia
    0.64
    oots
    0.63
    addle
    0.63
    rocal
    0.62
    fen
    0.61
    istered
    0.61
    Activations Density 0.025%

    No Known Activations