INDEX

Explanations

phrases related to food

oai_token-act-pair · gpt-3.5-turbo

terms and references related to food

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

 Wink

-0.89

-0.82

itudinal

-0.73

aird

-0.72

 Caldwell

-0.70

 Painter

-0.69

 Telegram

-0.68

Gib

-0.68

 Rowe

-0.66

 Werewolf

-0.65

POSITIVE LOGITS

stuff

1.14

borne

1.04

corn

1.01

 eaten

0.93

 additives

0.93

 poisoning

0.90

 allergies

0.90

 tasted

0.89

 starch

0.88

grain

0.86

Activations Density 0.026%

phrases related to food

terms and references related to food

No Comments

No Known Activations