© 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
    2. Apollo Research · Taylor · Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
    3. GPT2-Small
    4. Residual Similar CE E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 63
    Prev
    Next
    INDEX
    Explanations

    names of individuals, likely related to news, politics, or sports

    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
    92
    +0.04
    0.5%
    354
    +0.04
    0.5%
    606
    +0.04
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    556
    +0.04
    0.04
    406
    +0.04
    0.04
    373
    +0.04
    -0.02
    Negative Logits
    è¦ļéĨĴ
    -0.36
    ashtra
    -0.29
    EMP
    -0.29
    =-=-
    -0.28
    yahoo
    -0.27
    odynamic
    -0.27
    WARE
    -0.26
    erity
    -0.25
    ynthesis
    -0.24
    adelphia
    -0.24
    POSITIVE LOGITS
    theless
    0.31
    andom
    0.28
    ciating
    0.26
    issance
    0.26
    escent
    0.25
    isance
    0.25
     Cay
    0.24
    rence
    0.24
    ç
    0.24
    rences
    0.23
    Activations Density 0.258%

    No Known Activations