INDEX

Explanations

verbs indicating a response or reaction along with the word "react"

oai_token-act-pair · gpt-3.5-turbo

words related to reactions and responses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Fantastic

-0.71

fil

-0.70

lav

-0.66

thin

-0.65

scar

-0.65

 Twin

-0.63

ZA

-0.62

holding

-0.62

ammy

-0.62

 Mand

-0.60

POSITIVE LOGITS

 react

1.17

ivation

1.12

ivated

1.10

 reactions

1.04

ivating

0.96

uate

0.93

ively

0.93

 reacting

0.91

 reacts

0.90

uates

0.87

Activations Density 0.006%

verbs indicating a response or reaction along with the word "react"

words related to reactions and responses

No Comments

No Known Activations