© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 5569
    Prev
    Next
    INDEX
    Explanations

    high numbers, typically related to financial figures or statistics

    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
    +0.18
    0.6%
    639
    +0.17
    0.5%
    169
    +0.15
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    230
    +0.18
    0.06
    310
    +0.17
    0.06
    326
    +0.15
    0.05
    Negative Logits
    ulet
    -1.07
    bye
    -1.06
    infeld
    -1.04
     citiz
    -1.04
    emade
    -1.03
    idays
    -1.03
     fundament
    -1.02
    imen
    -1.00
    reon
    -0.98
     Bened
    -0.98
    POSITIVE LOGITS
    ths
    1.30
     lbs
    1.21
     pounds
    1.20
     ±
    1.19
     miles
    1.15
     yuan
    1.14
    ±
    1.14
     sq
    1.11
    AUD
    1.10
     yards
    1.08
    Activations Density 0.596%

    No Known Activations