INDEX

Explanations

text pieces indicating image captions with a focus on medical topics

oai_token-act-pair · gpt-3.5-turbo

image captions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

onest

-0.85

 reborn

-0.72

aturdays

-0.70

 empt

-0.69

fleet

-0.66

etheless

-0.66

eday

-0.66

glas

-0.65

quart

-0.65

avorite

-0.65

POSITIVE LOGITS

UTERS

0.87

 Provided

0.83

 Immun

0.77

=>

0.74

 Surveillance

0.73

 Image

0.72

Streamer

0.72

 IMAGES

0.71

 Protesters

0.70

 Researchers

0.70

Activations Density 0.065%

text pieces indicating image captions with a focus on medical topics

image captions

No Comments

No Known Activations