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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 45227
    Prev
    Next
    INDEX
    Explanations

    detail-oriented words and information, including names and specific events

    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
    421
    +0.15
    0.7%
    447
    +0.15
    0.7%
    57
    +0.15
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    57
    +0.15
    0.08
    75
    +0.15
    0.09
    421
    +0.15
    0.08
    Negative Logits
    DragonMagazine
    -0.81
    æĸ¹
    -0.73
     Citizen
    -0.71
    gerald
    -0.70
     Presidential
    -0.69
    ordinate
    -0.64
     Extras
    -0.63
    IFIED
    -0.63
     Cherokee
    -0.62
    æĺ¯
    -0.61
    POSITIVE LOGITS
    atonin
    1.23
    inda
    1.20
    bourne
    1.04
    aser
    0.99
    ting
    0.95
    issa
    0.95
    anches
    0.93
    odies
    0.93
    odic
    0.93
    ice
    0.92
    Activations Density 6.334%

    No Known Activations