INDEX

Explanations

references to breaks or interruptions in various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "break."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

itatively

-0.77

ifice

-0.76

Reviewer

-0.70

ãĥĺãĥ©

-0.68

 Cosponsors

-0.65

IFIED

-0.65

 obser

-0.62

oka

-0.60

alez

-0.59

 portrayal

-0.58

POSITIVE LOGITS

away

1.19

neck

1.10

fast

1.01

points

0.93

water

0.91

down

0.89

downs

0.89

aways

0.88

robe

0.87

breakers

0.83

Activations Density 0.036%

references to breaks or interruptions in various contexts

instances of the word "break."

No Comments

No Known Activations