© 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. 11-RES-JB
    11. 17636
    Prev
    Next
    INDEX
    Explanations

    words that express surprise or opposition

    oai_token-act-pair · gpt-3.5-turbo

    phrases that express uncertainty or the subjective perception of ideas or situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Surv
    -0.78
    ESE
    -0.76
    "]=>
    -0.75
     Returning
    -0.75
     Preferred
    -0.74
    urga
    -0.74
    cised
    -0.73
    essor
    -0.72
     Registrar
    -0.67
    ickr
    -0.66
    POSITIVE LOGITS
     daunting
    1.29
     trivial
    1.29
     insignificant
    1.22
     frivolous
    1.20
     outlandish
    1.19
     contradictory
    1.18
     cliché
    1.18
     innocuous
    1.16
     esoteric
    1.15
     paradox
    1.14
    Activations Density 0.090%

    No Known Activations