© 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. 4-RES-JB
    11. 22342
    Prev
    Next
    INDEX
    Explanations

    phrases related to explanations or justifications

    oai_token-act-pair · gpt-3.5-turbo

    terms related to reasoning and rationale

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

    No Comments

    Negative Logits
    chance
    -0.74
    semble
    -0.73
    hold
    -0.69
    obl
    -0.68
    uden
    -0.68
    touch
    -0.67
    adal
    -0.67
    tein
    -0.67
    bane
    -0.66
    oner
    -0.66
    POSITIVE LOGITS
     rationale
    1.08
    SourceFile
    1.03
     why
    1.02
     justification
    0.98
     reasoning
    0.97
     WHY
    0.87
     justifying
    0.87
     behind
    0.86
     argument
    0.83
     underpin
    0.80
    Activations Density 0.018%

    No Known Activations