© 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. 1393
    Prev
    Next
    INDEX
    Explanations

    references to software, technology, and cryptocurrency-related terms, including the name "BitDefender"

    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
    715
    +0.15
    0.5%
    79
    +0.14
    0.5%
    589
    +0.14
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    30
    +0.15
    0.03
    715
    +0.14
    0.04
    13
    +0.14
    0.04
    Negative Logits
    ammed
    -1.13
    ammad
    -1.09
    gaard
    -0.96
    ordan
    -0.89
    ksh
    -0.89
    erved
    -0.88
    qqa
    -0.87
    etheus
    -0.84
     polarized
    -0.83
    angler
    -0.83
    POSITIVE LOGITS
    terness
    1.53
    coins
    1.43
    buck
    1.32
    maps
    1.23
    umen
    1.17
    fine
    1.13
    ches
    1.11
    lass
    1.07
    rarily
    1.05
    map
    1.04
    Activations Density 0.653%

    No Known Activations