© 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
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE Local
    9. 6-RES_SCL-AJT
    11. 40308
    Prev
    Next
    INDEX
    Explanations

    measurements in kilograms and pounds

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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
    23
    +0.13
    0.6%
    255
    +0.12
    0.6%
    326
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    535
    +0.13
    0.05
    23
    +0.12
    0.09
    176
    +0.11
    0.05
    Negative Logits
    olitan
    -0.77
    eering
    -0.76
    itual
    -0.73
     VK
    -0.70
    idate
    -0.70
    western
    -0.69
    idates
    -0.64
    actionDate
    -0.64
    itized
    -0.64
    ãĥ¤
    -0.64
    POSITIVE LOGITS
    hered
    0.78
    airs
    0.76
    lbs
    0.75
    lb
    0.75
    \\\\\\\\
    0.74
    ubb
    0.72
     lbs
    0.71
    reat
    0.70
    icka
    0.69
    eki
    0.69
    Activations Density 8.846%

    No Known Activations