© 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
    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. 14287
    Prev
    Next
    INDEX
    Explanations

    references to legal proceedings and investigations

    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
    +0.24
    1.1%
    83
    +0.14
    0.6%
    393
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.24
    0.39
    640
    +0.14
    0.37
    739
    +0.11
    0.07
    Negative Logits
    nikov
    -0.76
    angelo
    -0.72
    ilus
    -0.71
    osaurus
    -0.66
     Sands
    -0.66
    lie
    -0.65
    ãĥ¼
    -0.63
    rise
    -0.62
    lust
    -0.62
    isky
    -0.62
    POSITIVE LOGITS
    ASY
    0.88
    HAM
    0.86
    CI
    0.85
    ITION
    0.85
    NET
    0.82
     acronym
    0.81
    COM
    0.81
    RD
    0.80
    NA
    0.79
    PS
    0.79
    Activations Density 21.399%

    No Known Activations