INDEX

Explanations

phrases containing the word "no" with varying intensity

oai_token-act-pair · gpt-3.5-turbo

repetitive phrases emphasizing the absence or lack of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

gypt

-0.78

rex

-0.65

 alot

-0.64

alian

-0.62

ortium

-0.62

adobe

-0.61

roxy

-0.61

romeda

-0.61

ËĪ

-0.60

nox

-0.60

POSITIVE LOGITS

xious

1.13

except

1.04

 discern

0.94

 such

0.88

oses

0.86

 indication

0.86

 illusions

0.84

 shortage

0.84

 qual

0.83

Such

0.82

Activations Density 0.038%

phrases containing the word "no" with varying intensity

repetitive phrases emphasizing the absence or lack of something

No Comments

No Known Activations