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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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 E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 212
    Prev
    Next
    INDEX
    Explanations

    references to physical body parts and physical activities related to exercise and movement

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scefr-ajt/6-res_scefr-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
    709
    +0.16
    0.5%
    624
    +0.16
    0.5%
    754
    +0.15
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    99
    +0.16
    0.04
    754
    +0.16
    0.04
    345
    +0.15
    0.04
    Negative Logits
     Kafka
    -1.16
     Hoover
    -1.03
    Ń·
    -0.94
    yrinth
    -0.91
     Tart
    -0.90
    urations
    -0.89
    ãĥĻ
    -0.89
     Fundamental
    -0.89
    TAG
    -0.88
     Dickens
    -0.88
    POSITIVE LOGITS
    guards
    1.91
    guard
    1.69
    builders
    1.47
    building
    1.42
    anguage
    1.38
    builder
    1.36
     politic
    1.32
    weight
    1.24
    parts
    1.23
    wash
    1.23
    Activations Density 0.501%

    No Known Activations