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. 13918
    Prev
    Next
    INDEX
    Explanations

    mentions of specific numbers or statistics in the context of various topics

    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
    288
    +0.16
    0.6%
    575
    +0.13
    0.5%
    711
    +0.12
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.16
    0.19
    288
    +0.13
    0.19
    225
    +0.12
    0.12
    Negative Logits
    nick
    -0.82
     Trin
    -0.81
     pree
    -0.74
     Tru
    -0.74
    established
    -0.72
    going
    -0.70
     Ly
    -0.70
     Po
    -0.70
     Rail
    -0.70
    Beg
    -0.69
    POSITIVE LOGITS
     abroad
    1.24
     indoors
    1.16
     whilst
    1.13
     accordingly
    1.08
     outdoors
    1.06
     onstage
    1.05
    poons
    1.00
    enium
    0.99
     WITHOUT
    0.99
    20439
    0.98
    Activations Density 3.051%

    No Known Activations