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 Local
    5. 6-RES_SCL-AJT
    6. 34499
    Prev
    Next
    INDEX
    Explanations

    phrases related to long-term projects or initiatives

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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
    497
    +0.05
    0.4%
    317
    +0.05
    0.4%
    540
    +0.04
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.05
    0.17
    575
    +0.05
    0.13
    651
    +0.04
    0.10
    Negative Logits
    cop
    -0.41
     cop
    -0.41
     stret
    -0.41
    ©¶æ
    -0.38
    pat
    -0.38
     exting
    -0.36
     mum
    -0.35
    oping
    -0.34
     widow
    -0.34
     responders
    -0.34
    POSITIVE LOGITS
    ĸļ
    0.44
     largeDownload
    0.41
    DragonMagazine
    0.38
    Craft
    0.37
    UID
    0.36
    ZI
    0.36
     guiActiveUnfocused
    0.36
    Reward
    0.36
    ADE
    0.35
    Forge
    0.34
    Activations Density 4.032%

    No Known Activations