Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 13819
    Prev
    Next
    INDEX
    Explanations

    references to specific geographical locations, particularly countries such as Cameroon and Nigeria

    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
    461
    +0.07
    0.5%
    623
    +0.07
    0.5%
    10
    +0.06
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.07
    0.02
    643
    +0.07
    0.01
    238
    +0.06
    0.01
    Negative Logits
    PDATE
    -0.46
    VK
    -0.43
    fts
    -0.42
    Eat
    -0.42
    HY
    -0.41
    heng
    -0.39
    EEE
    -0.39
    cor
    -0.38
    aepernick
    -0.38
    Ô
    -0.38
    POSITIVE LOGITS
    backs
    0.45
    alo
    0.44
    ians
    0.44
    icans
    0.43
    ascript
    0.42
     Ou
    0.42
     Sturgeon
    0.42
    nil
    0.42
    efully
    0.40
    upon
    0.40
    Activations Density 0.061%

    No Known Activations