INDEX

Explanations

mentions of ingredients or steps in a recipe that involve chopping

oai_token-act-pair · gpt-3.5-turbo

instances of the word "chopped" and its variations in cooking contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ports

-0.76

forcing

-0.73

porting

-0.73

isle

-0.73

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.71

Grey

-0.68

ships

-0.68

 Gutenberg

-0.66

SPONSORED

-0.66

VERTIS

-0.66

POSITIVE LOGITS

 chop

1.17

 Chop

0.94

 onions

0.93

 chunks

0.92

 chops

0.90

 slices

0.89

aneous

0.89

sticks

0.88

 onion

0.85

 chopping

0.83

Activations Density 0.020%

mentions of ingredients or steps in a recipe that involve chopping

instances of the word "chopped" and its variations in cooking contexts

No Comments

No Known Activations