© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 12288
    5. 8-RES_FS12288-JB
    6. 1682
    Prev
    Next
    INDEX
    Explanations

    text related to verifying human identity online

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    125
    +0.10
    0.5%
    208
    +0.10
    0.4%
    257
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    521
    +0.10
    0.02
    208
    +0.10
    0.02
    125
    +0.08
    0.02
    Negative Logits
     prototyp
    -0.68
    pmwiki
    -0.65
    NetMessage
    -0.63
     unaccount
    -0.61
     brainstorm
    -0.61
    ibaba
    -0.59
     tantal
    -0.59
     Guid
    -0.58
     polyg
    -0.58
    aimon
    -0.58
    POSITIVE LOGITS
    email
    0.76
    iframe
    0.68
    api
    0.67
    VPN
    0.63
    $$$$
    0.62
    RS
    0.62
    smoking
    0.60
    usha
    0.60
     php
    0.60
    cens
    0.60
    Activations Density 0.046%

    No Known Activations