© 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. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 1759
    Prev
    Next
    INDEX
    Explanations

    mentions of physical actions and bodily sensations

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-ajt
    Prompts (Dashboard)
    12,288 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    46,080
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    apollo-research/Skylion007-openwebtext-tokenizer-gpt2
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    177
    +0.09
    0.4%
    398
    +0.09
    0.4%
    526
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    188
    +0.09
    0.04
    354
    +0.09
    0.04
    128
    +0.08
    0.04
    Negative Logits
    £ı
    -0.81
    Ħ¢
    -0.63
    Ĥª
    -0.62
    ĨĴ
    -0.61
    clerosis
    -0.61
    reciation
    -0.60
    iliary
    -0.59
    ashtra
    -0.58
    BIL
    -0.57
     Letters
    -0.57
    POSITIVE LOGITS
    nit
    0.72
    oused
    0.69
    ered
    0.69
    mouth
    0.64
    gered
    0.63
    ogging
    0.62
    chest
    0.61
    ike
    0.61
    itted
    0.61
    ashing
    0.60
    Activations Density 0.257%

    No Known Activations