INDEX

Explanations

statements related to potential health risks or dangers

oai_token-act-pair · gpt-3.5-turbo

statements describing the characteristics or categorizations of substances

oai_token-act-pair · gpt-4o-mini Triggered 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

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

onnaissance

-0.80

isions

-0.78

 countdown

-0.70

ision

-0.70

 Pictures

-0.69

 Isles

-0.68

endars

-0.68

 Checking

-0.68

Kings

-0.67

edin

-0.67

POSITIVE LOGITS

 marketed

1.25

 chemically

1.19

 soluble

1.19

 ingested

1.16

 metabol

1.12

 abundant

1.11

 harvested

1.10

 sprayed

1.09

 commonly

1.09

 readily

1.09

Activations Density 0.176%

statements related to potential health risks or dangers

statements describing the characteristics or categorizations of substances

No Comments

No Known Activations