© 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. 2
    Prev
    Next
    INDEX
    Explanations

    the word "don't" followed by a verb suggesting a prohibition or warning

    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
    326
    +0.13
    0.5%
    360
    +0.13
    0.5%
    669
    +0.12
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    360
    +0.13
    0.06
    61
    +0.13
    0.05
    643
    +0.12
    0.04
    Negative Logits
    ELD
    -0.99
    milo
    -0.96
    creen
    -0.86
    gnu
    -0.85
    edge
    -0.84
    chrom
    -0.84
    potion
    -0.83
    elvet
    -0.83
    CVE
    -0.82
    cream
    -0.82
    POSITIVE LOGITS
     hesitate
    1.08
     forget
    1.02
     yourselves
    0.96
     worry
    0.96
     underestimate
    0.95
     confuse
    0.90
     kidding
    0.88
     thou
    0.87
     yourself
    0.86
     bother
    0.85
    Activations Density 0.479%

    No Known Activations