INDEX

Explanations

adjectives describing a negative or worrying situation

oai_token-act-pair · gpt-3.5-turbo

terms indicating danger or concern

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

neys

-1.02

pel

-0.98

PB

-0.87

tein

-0.84

bid

-0.84

nan

-0.83

ney

-0.81

dain

-0.81

angled

-0.81

ordan

-0.79

POSITIVE LOGITS

 alarming

0.96

 mosqu

0.88

 Cabrera

0.88

 omin

0.88

 undermin

0.84

 suspic

0.79

 ominous

0.78

 alarm

0.78

 warnings

0.77

ÃĽ

0.75

Activations Density 0.014%

adjectives describing a negative or worrying situation

terms indicating danger or concern

No Comments

No Known Activations