© 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. 1467
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to individuals and locations

    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
    679
    +0.17
    0.6%
    765
    +0.14
    0.5%
    708
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    101
    +0.17
    0.10
    603
    +0.14
    0.06
    725
    +0.13
    0.06
    Negative Logits
    agos
    -0.98
    qua
    -0.98
    ework
    -0.97
    irth
    -0.91
    sth
    -0.90
    PLIED
    -0.89
    stead
    -0.86
    UTERS
    -0.85
    issance
    -0.85
    ivation
    -0.83
    POSITIVE LOGITS
    paces
    1.14
    essee
    1.08
     names
    1.07
     aliases
    0.97
     spots
    0.94
    lime
    0.93
     landmarks
    0.82
     scapego
    0.82
    onyms
    0.82
     keywords
    0.82
    Activations Density 1.502%

    No Known Activations