© 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. 13633
    Prev
    Next
    INDEX
    Explanations

    character names from a story or text

    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
    326
    +0.15
    0.6%
    167
    +0.14
    0.5%
    14
    +0.14
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.15
    0.08
    288
    +0.14
    0.09
    542
    +0.14
    0.05
    Negative Logits
    ktop
    -1.10
    dit
    -1.00
    SPONSORED
    -0.94
    iator
    -0.93
    iour
    -0.91
    cha
    -0.90
    iox
    -0.88
     Gareth
    -0.88
    olate
    -0.87
    erness
    -0.82
    POSITIVE LOGITS
    Ruby
    1.04
     gems
    0.92
    BY
    0.90
    gem
    0.90
     Ruby
    0.90
     Flask
    0.90
    aclysm
    0.89
     Carbuncle
    0.89
     Gems
    0.88
    claw
    0.86
    Activations Density 1.274%

    No Known Activations