© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    5. GPT2-Small
    7. Residual Similar CE E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 11961
    Prev
    Next
    INDEX
    Explanations

    sports-related terms, specifically related to football positions and events

    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
    0
    +0.16
    0.6%
    60
    +0.14
    0.5%
    276
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    0
    +0.16
    0.06
    749
    +0.14
    0.05
    60
    +0.13
    0.04
    Negative Logits
    owship
    -1.02
    ãĤ¼ãĤ¦ãĤ¹
    -0.93
     straw
    -0.88
     Dise
    -0.86
    NESS
    -0.86
    ibrary
    -0.85
    worms
    -0.83
    ource
    -0.82
     halls
    -0.81
     Graveyard
    -0.80
    POSITIVE LOGITS
    digy
    2.03
    verbs
    1.60
    gression
    1.53
    blems
    1.45
    posal
    1.41
    secution
    1.35
    clamation
    1.34
    ctor
    1.33
    ccess
    1.30
    ceed
    1.28
    Activations Density 0.667%

    No Known Activations