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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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 Local
    5. 6-RES_SCL-AJT
    6. 34548
    Prev
    Next
    INDEX
    Explanations

    declassified classified information or documents

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-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
    447
    +0.15
    0.9%
    111
    +0.08
    0.5%
    481
    +0.07
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    479
    +0.15
    0.03
    640
    +0.08
    0.21
    301
    +0.07
    0.08
    Negative Logits
    owitz
    -0.64
    osity
    -0.62
    Goal
    -0.54
    ager
    -0.53
    apo
    -0.52
    oult
    -0.51
    asso
    -0.51
    ptoms
    -0.50
    idity
    -0.50
    abl
    -0.50
    POSITIVE LOGITS
     documents
    0.58
     copies
    0.58
    nesday
    0.58
     transcripts
    0.57
     manuscript
    0.57
     Territories
    0.55
     manuscripts
    0.55
    foundland
    0.53
     versions
    0.53
     Kenyan
    0.52
    Activations Density 10.187%

    No Known Activations