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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 40347
    Prev
    Next
    INDEX
    Explanations

    technical information related to computer architecture and testing procedures

    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
    417
    +0.10
    0.5%
    678
    +0.10
    0.5%
    11
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    99
    +0.10
    0.11
    326
    +0.10
    0.21
    137
    +0.10
    0.05
    Negative Logits
    ĸļ
    -0.73
     Veil
    -0.66
    inus
    -0.63
    acus
    -0.63
     mete
    -0.63
     Hutch
    -0.61
     Faust
    -0.61
     Rasmussen
    -0.60
    ĪĴ
    -0.58
     shelter
    -0.57
    POSITIVE LOGITS
    DD
    0.71
    Disk
    0.70
    ODE
    0.68
     gaming
    0.67
     GPUs
    0.64
     chars
    0.64
     encoded
    0.63
    WARE
    0.63
    BSD
    0.63
    Gaming
    0.62
    Activations Density 8.518%

    No Known Activations