Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 418
    Prev
    Next
    INDEX
    Explanations

    code syntax and technical keywords within a software development context

    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
    238
    +0.03
    0.4%
    438
    +0.03
    0.4%
    526
    +0.03
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    424
    +0.03
    0.01
    326
    +0.03
    0.01
    328
    +0.03
    0.00
    Negative Logits
     Encyclopedia
    -0.21
     Explore
    -0.20
     THR
    -0.19
     Online
    -0.19
     Archive
    -0.19
    ulu
    -0.19
    illin
    -0.19
     Vanguard
    -0.18
     å¤
    -0.18
    efully
    -0.18
    POSITIVE LOGITS
    ©¶æ
    0.28
     pse
    0.22
    idis
    0.21
    mberg
    0.21
     seams
    0.21
    gged
    0.21
    headed
    0.20
     psychiat
    0.20
    velt
    0.20
     misled
    0.20
    Activations Density 0.022%

    No Known Activations