INDEX

Explanations

captions from images with the word "toggle"

oai_token-act-pair · gpt-3.5-turbo

commands related to image interaction or display options

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

bats

-0.71

ivism

-0.64

comed

-0.63

 Expend

-0.61

stood

-0.60

dies

-0.60

warm

-0.58

rency

-0.58

 hindsight

-0.58

naires

-0.57

POSITIVE LOGITS

 caption

1.24

 toggle

0.85

taboola

0.82

ption

0.79

%]

0.75

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

0.69

 copyright

0.69

 Reuters

0.66

 Courtesy

0.66

ONSORED

0.66

Activations Density 0.008%

captions from images with the word "toggle"

commands related to image interaction or display options

No Comments

No Known Activations