INDEX

Explanations

image captions in a document

oai_token-act-pair · gpt-3.5-turbo

segments related to image captions

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

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

limited

-0.65

urity

-0.64

aea

-0.62

 Relic

-0.62

fir

-0.59

Gy

-0.59

Tom

-0.58

ste

-0.58

nut

-0.58

iable

-0.57

POSITIVE LOGITS

 caption

4.40

 Caption

1.61

 capt

1.46

 copyright

1.41

 subtitle

1.34

 subtitles

1.25

 slideshow

1.12

 annotations

1.10

 hasht

1.07

 narrator

1.07

Activations Density 0.019%

image captions in a document

segments related to image captions

No Comments

No Known Activations