© 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. 10-RES-JB
    11. 12659
    Prev
    Next
    INDEX
    Explanations

    verbs followed by the infinitive "to" as part of an expression of intention or desire

    oai_token-act-pair · gpt-3.5-turbo

    phrases about attempts, desires, and willingness related to actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     assurances
    -0.75
    Þ
    -0.72
    ij士
    -0.72
     reassuring
    -0.70
     weakening
    -0.69
    İĭ
    -0.68
     vanishing
    -0.68
     overriding
    -0.67
    terness
    -0.66
     deterrence
    -0.66
    POSITIVE LOGITS
     participate
    1.29
     attend
    1.22
     partake
    1.17
     hear
    1.12
     visit
    1.09
     listen
    1.07
     watch
    1.00
     bother
    0.99
     try
    0.99
     join
    0.99
    Activations Density 0.318%

    No Known Activations