Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 21920
    Prev
    Next
    INDEX
    Explanations

    references to a specific physical object - a barrel

    oai_token-act-pair · gpt-3.5-turbo

    references to barrels in relation to quantities and pricing

    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
    ãħĭãħĭ
    -0.76
    ãħĭ
    -0.76
    android
    -0.70
     Conscious
    -0.67
    ults
    -0.66
    iza
    -0.65
     Scientology
    -0.65
    èª
    -0.64
    Parents
    -0.64
    cong
    -0.63
    POSITIVE LOGITS
     barrel
    3.91
     barrels
    3.01
     Barrel
    2.87
     muzzle
    1.36
     cellar
    1.34
     bottle
    1.22
     piston
    1.19
     cannon
    1.15
     turret
    1.13
     blade
    1.11
    Activations Density 0.027%

    No Known Activations