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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 3725
    Prev
    Next
    INDEX
    Explanations

    terms related to advocacy and support for specific causes or movements

    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
    679
    +0.29
    1.1%
    288
    +0.23
    0.9%
    644
    +0.17
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.29
    0.12
    247
    +0.23
    0.09
    679
    +0.17
    0.11
    Negative Logits
    kered
    -0.83
    obyl
    -0.80
    plet
    -0.80
     gorge
    -0.79
     arsen
    -0.79
    uton
    -0.76
    peg
    -0.74
    oba
    -0.74
    âĢ¢âĢ¢âĢ¢âĢ¢
    -0.73
    Lay
    -0.73
    POSITIVE LOGITS
     Pwr
    0.91
    network
    0.89
     advocating
    0.89
    leading
    0.87
    ersed
    0.85
    leader
    0.85
    orer
    0.83
    iguous
    0.83
    tip
    0.80
    byss
    0.78
    Activations Density 1.604%

    No Known Activations