© 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 Local
    9. 6-RES_SCL-AJT
    11. 3988
    Prev
    Next
    INDEX
    Explanations

    pros and cons listed in a review or evaluation text related to food and restaurants

    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
    420
    +0.03
    0.5%
    389
    +0.03
    0.5%
    156
    +0.03
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    27
    +0.03
    0.05
    285
    +0.03
    0.04
    311
    +0.03
    -0.04
    Negative Logits
    ĸļ
    -0.21
     AUTHOR
    -0.20
    arov
    -0.20
     stret
    -0.20
    CHAT
    -0.19
    DragonMagazine
    -0.18
     TRAN
    -0.18
    20439
    -0.18
    Blocks
    -0.18
     retrie
    -0.17
    POSITIVE LOGITS
    utral
    0.22
    estine
    0.20
    luster
    0.20
    vertisements
    0.19
    itu
    0.19
    advant
    0.18
     corro
    0.18
    sters
    0.17
    ÙĴ
    0.17
    apple
    0.17
    Activations Density 5.182%

    No Known Activations