© 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. 14027
    Prev
    Next
    INDEX
    Explanations

    numeric patterns and numerical values such as specific numbers or numerical expressions

    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
    383
    +0.17
    0.7%
    374
    +0.15
    0.6%
    166
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    740
    +0.17
    0.07
    612
    +0.15
    0.06
    166
    +0.13
    0.06
    Negative Logits
    loo
    -0.95
    WARD
    -0.76
    EEE
    -0.73
    REDACTED
    -0.72
    hire
    -0.71
    mosp
    -0.71
    ELF
    -0.71
     Template
    -0.70
    hold
    -0.70
     Directive
    -0.69
    POSITIVE LOGITS
    eral
    1.02
    pty
    0.96
    BER
    0.94
    emonic
    0.92
    ptoms
    0.91
    itionally
    0.90
    phys
    0.90
    locked
    0.89
    ming
    0.88
     num
    0.84
    Activations Density 1.153%

    No Known Activations