© 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
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 1-RES-JB
    11. 16162
    Prev
    Next
    INDEX
    Explanations

    phrases mentioning contrasts or conflicts

    oai_token-act-pair · gpt-3.5-turbo

    strong punctuation usage, particularly commas

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     corrid
    -0.60
     referen
    -0.51
     predec
    -0.50
    thora
    -0.48
    igmatic
    -0.44
    estones
    -0.44
     thous
    -0.43
    arij
    -0.43
    ãĥ¼ãĥĨãĤ£
    -0.42
    ¬¼
    -0.41
    POSITIVE LOGITS
     albeit
    0.59
     excluding
    0.53
     meanwhile
    0.51
     namely
    0.50
     huh
    0.50
     although
    0.50
     Kinnikuman
    0.49
    apa
    0.49
    ktop
    0.48
    please
    0.47
    Activations Density 0.825%

    No Known Activations