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. 34665
    Prev
    Next
    INDEX
    Explanations

    information related to technology, television, and events such as the Olympics and Oscars

    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
    +0.26
    1.3%
    481
    +0.26
    1.3%
    459
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    481
    +0.26
    0.32
    459
    +0.26
    0.17
    482
    +0.11
    0.08
    Negative Logits
     Sutherland
    -0.69
    SY
    -0.66
    bil
    -0.65
    dfx
    -0.65
    lev
    -0.63
    ãĥ¼ãĥ
    -0.62
    lip
    -0.62
    stadt
    -0.61
    470
    -0.61
    bour
    -0.61
    POSITIVE LOGITS
     anniversary
    0.72
    iversary
    0.72
     festivities
    0.72
     outcome
    0.70
     voyage
    0.70
     lasts
    0.70
     season
    0.69
     progresses
    0.68
     finale
    0.68
     date
    0.68
    Activations Density 13.097%

    No Known Activations