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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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 Local
    9. 6-RES_SCL-AJT
    11. 13887
    Prev
    Next
    INDEX
    Explanations

    terms related to competitive matchups or contests

    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
    360
    +0.07
    0.4%
    566
    +0.07
    0.4%
    610
    +0.07
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.07
    0.22
    326
    +0.07
    0.23
    541
    +0.07
    0.10
    Negative Logits
    zynski
    -0.64
    authorized
    -0.61
    ruption
    -0.58
    alties
    -0.58
    ifted
    -0.58
    mington
    -0.56
    oup
    -0.56
    sonian
    -0.55
    lear
    -0.55
    isks
    -0.55
    POSITIVE LOGITS
     matchups
    0.63
    à©
    0.57
     inacc
    0.55
    agame
    0.54
     matchup
    0.54
     ailments
    0.53
    à¨
    0.52
     nightmares
    0.49
     Pv
    0.48
     sidew
    0.48
    Activations Density 8.753%

    No Known Activations