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. 4528
    Prev
    Next
    INDEX
    Explanations

    names of famous people, political entities, and contentious topics

    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
    +0.17
    0.6%
    187
    +0.15
    0.6%
    393
    +0.15
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.17
    0.23
    393
    +0.15
    0.26
    270
    +0.15
    0.04
    Negative Logits
    assets
    -0.86
    effic
    -0.85
    gran
    -0.83
     proportions
    -0.79
    tips
    -0.75
     vulnerabilities
    -0.72
    efficiency
    -0.71
    generated
    -0.70
    nutrition
    -0.70
    ÅĤ
    -0.70
    POSITIVE LOGITS
     fray
    1.75
     bandwagon
    1.25
     chorus
    1.20
     ranks
    1.01
     fellowship
    0.89
     fold
    0.89
    rocal
    0.81
     conversation
    0.80
    CF
    0.80
    neau
    0.79
    Activations Density 12.719%

    No Known Activations