© 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. 5301
    Prev
    Next
    INDEX
    Explanations

    statements expressing certainty or belief

    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
    393
    +0.34
    1.1%
    102
    +0.18
    0.6%
    208
    +0.16
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.34
    0.13
    102
    +0.18
    0.14
    566
    +0.16
    0.10
    Negative Logits
    anwhile
    -1.10
    rones
    -1.01
     Capitals
    -0.94
    osate
    -0.94
    ossom
    -0.92
    avanaugh
    -0.91
    atures
    -0.90
     Mund
    -0.90
     Sebast
    -0.89
    plings
    -0.87
    POSITIVE LOGITS
     able
    1.48
    leeve
    1.37
     considered
    1.21
     regarded
    1.21
     quite
    1.20
    fits
    1.20
    que
    1.18
     phased
    1.16
     influenced
    1.15
    hemoth
    1.15
    Activations Density 1.467%

    No Known Activations