© 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 E2E Recon
    9. 6-RES_SCEFR-AJT
    11. 14157
    Prev
    Next
    INDEX
    Explanations

    traffic stops and interactions with law enforcement

    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
    104
    +0.15
    0.5%
    538
    +0.13
    0.4%
    496
    +0.13
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.15
    0.09
    326
    +0.13
    0.07
    70
    +0.13
    0.06
    Negative Logits
    wcs
    -1.25
    cffff
    -1.19
    rina
    -1.18
    yss
    -1.17
    idia
    -1.12
    ochond
    -1.09
    sonian
    -1.08
    inth
    -1.08
    ricanes
    -1.06
    女
    -1.05
    POSITIVE LOGITS
     Glock
    1.20
     checkpoint
    1.17
     motorists
    1.16
     checkpoints
    1.14
     DEA
    1.11
     parked
    1.08
     parking
    1.08
     BART
    1.08
     patrol
    1.07
     cops
    1.06
    Activations Density 0.903%

    No Known Activations