INDEX

Explanations

words related to rodents, particularly rats

oai_token-act-pair · gpt-3.5-turbo

references to rats and related rodent terminology

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

verson

-0.74

gur

-0.72

othing

-0.72

ctic

-0.72

atever

-0.70

olulu

-0.65

planned

-0.65

otide

-0.64

 Sikh

-0.64

vot

-0.62

POSITIVE LOGITS

 Rats

1.08

 rats

1.03

lings

1.02

rats

0.88

 rodents

0.86

folk

0.84

 beetles

0.79

dayName

0.79

 mice

0.77

MpServer

0.77

Activations Density 0.009%

words related to rodents, particularly rats

references to rats and related rodent terminology

No Comments

No Known Activations