© 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. 8-RES-JB
    11. 1761
    Prev
    Next
    INDEX
    Explanations

    descriptions of physical attributes or characteristics

    oai_token-act-pair · gpt-3.5-turbo

    instances of heightened emotional or physical sensations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    asions
    -0.59
     were
    -0.59
    yss
    -0.58
    iances
    -0.56
     convened
    -0.56
     promotions
    -0.55
    artments
    -0.53
     Promotion
    -0.53
    arij
    -0.52
     specialize
    -0.52
    POSITIVE LOGITS
    ?).
    0.70
    ?".
    0.67
    vironment
    0.65
     itself
    0.65
    Redditor
    0.65
     its
    0.63
    etheless
    0.61
    pron
    0.60
    Ö
    0.59
    )?
    0.59
    Activations Density 1.503%

    No Known Activations