© 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. 22301
    Prev
    Next
    INDEX
    Explanations

    specifications or mentions of technology models with the 'X' identifier

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the letter 'X' in various contexts

    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
     Beth
    -0.70
     Mol
    -0.70
     gut
    -0.66
     deliber
    -0.64
     sustaining
    -0.64
     sustained
    -0.64
     constituency
    -0.64
     Ber
    -0.63
     celebr
    -0.63
     bri
    -0.62
    POSITIVE LOGITS
    X
    3.88
    XX
    2.08
    XL
    2.02
     X
    2.01
    x
    1.99
    XT
    1.95
    XP
    1.95
    XM
    1.78
    Xi
    1.76
    XXX
    1.75
    Activations Density 0.021%

    No Known Activations