© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 0-RES-JB
    11. 11547
    Prev
    Next
    INDEX
    Explanations

    references to margins or differences in value or votes

    oai_token-act-pair · gpt-3.5-turbo

    references to margins, particularly in contexts involving measurements or comparisons

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Therapy
    -0.67
    hen
    -0.63
    Det
    -0.62
    hun
    -0.61
    ochet
    -0.61
    Aid
    -0.61
    activ
    -0.61
    det
    -0.61
    Demon
    -0.60
     worsh
    -0.59
    POSITIVE LOGITS
     margin
    4.18
     margins
    3.33
    margin
    2.70
    marg
    1.49
    Marg
    1.38
     Marg
    1.37
     cushion
    1.11
     edge
    1.10
    padding
    1.07
     ratio
    1.04
    Activations Density 0.018%

    No Known Activations