INDEX

Explanations

mentions of different types of candies

oai_token-act-pair · gpt-3.5-turbo

references to candy

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

yon

-0.69

lihood

-0.68

inen

-0.66

productive

-0.65

Published

-0.64

chron

-0.62

heed

-0.61

iets

-0.61

 transcript

-0.61

 Hutchinson

-0.61

POSITIVE LOGITS

 cane

1.12

 candy

0.94

strip

0.93

mallow

0.92

flake

0.88

wra

0.85

bucks

0.84

bar

0.83

 sweets

0.81

weet

0.81

Activations Density 0.032%

mentions of different types of candies

references to candy

No Comments

No Known Activations