© 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. 14076
    Prev
    Next
    INDEX
    Explanations

    specific numbers written as values

    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
    545
    +0.20
    0.6%
    163
    +0.18
    0.6%
    659
    +0.16
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    545
    +0.20
    0.07
    659
    +0.18
    0.06
    232
    +0.16
    0.06
    Negative Logits
    ĪĴ
    -1.25
    ¥µ
    -1.11
    ¿½
    -1.09
     millenn
    -1.06
    £ı
    -1.06
    anguage
    -1.02
     elbows
    -1.01
     bumper
    -1.00
     duplication
    -0.98
     mosqu
    -0.97
    POSITIVE LOGITS
    irus
    1.72
    ascular
    1.62
    endor
    1.61
    intage
    1.60
    olution
    1.59
    igil
    1.57
    apor
    1.56
    ampire
    1.53
    isions
    1.52
    ault
    1.50
    Activations Density 0.834%

    No Known Activations