© 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. 39950
    Prev
    Next
    INDEX
    Explanations

    descriptions of problems followed by an analysis or conclusion

    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
    447
    +0.30
    1.5%
    674
    +0.13
    0.6%
    393
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.30
    0.38
    14
    +0.13
    0.09
    739
    +0.09
    0.07
    Negative Logits
    throp
    -0.73
    ypes
    -0.70
    ervative
    -0.65
    chair
    -0.65
    emale
    -0.65
    chairs
    -0.64
    anmar
    -0.64
    opted
    -0.62
    notations
    -0.61
    asts
    -0.60
    POSITIVE LOGITS
     none
    0.73
     NOT
    0.68
     NEVER
    0.66
     ALWAYS
    0.65
     not
    0.64
     timing
    0.64
     compounded
    0.61
     Solitaire
    0.60
     neither
    0.57
     actually
    0.57
    Activations Density 18.300%

    No Known Activations