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 E2E Recon
    5. 6-RES_SCEFR-AJT
    6. 632
    Prev
    Next
    INDEX
    Explanations

    the keyword "paste"

    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
    321
    +0.09
    0.5%
    294
    +0.08
    0.5%
    239
    +0.07
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    321
    +0.09
    0.01
    239
    +0.08
    0.01
    203
    +0.07
    0.01
    Negative Logits
     Reserved
    -0.47
    bands
    -0.47
     blockade
    -0.46
    vernment
    -0.45
     viol
    -0.44
    FTWARE
    -0.44
     racket
    -0.43
     Portug
    -0.43
     conflic
    -0.43
     lifes
    -0.42
    POSITIVE LOGITS
    urized
    0.92
    ur
    0.73
    URES
    0.69
    urers
    0.65
     Paste
    0.65
    uring
    0.65
    ures
    0.63
    URE
    0.62
    urer
    0.57
    ilon
    0.56
    Activations Density 0.047%

    No Known Activations