Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE Local
    5. 6-RES_SCL-AJT
    6. 40286
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to legal and criminal activities

    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
    446
    +0.11
    0.5%
    649
    +0.10
    0.4%
    71
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    373
    +0.11
    -0.13
    231
    +0.10
    0.10
    276
    +0.09
    0.12
    Negative Logits
    stood
    -0.77
     gems
    -0.68
    lite
    -0.67
     Haku
    -0.67
    )",
    -0.66
     tongues
    -0.65
    itiz
    -0.65
     ().
    -0.63
    ties
    -0.63
     colours
    -0.63
    POSITIVE LOGITS
     WATCHED
    1.15
     Expand
    0.95
    Photos
    0.95
    Correction
    0.94
     Continued
    0.93
    Read
    0.92
     UNCLASSIFIED
    0.92
     SHARES
    0.90
     Recap
    0.87
    Follow
    0.87
    Activations Density 5.301%

    No Known Activations