INDEX

Explanations

words related to opposition or resistance

oai_token-act-pair · gpt-3.5-turbo

words related to assumptions or beliefs

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SEA

-0.85

 Masters

-0.83

AMS

-0.83

GER

-0.80

 Timber

-0.74

DER

-0.73

 Abyssal

-0.72

Forest

-0.72

 Schneider

-0.72

 Rite

-0.71

POSITIVE LOGITS

 supp

1.30

 scrut

1.05

lication

0.99

osition

0.95

ressive

0.94

pse

0.92

orter

0.90

roleum

0.86

uration

0.85

 plaus

0.84

Activations Density 0.008%

words related to opposition or resistance

words related to assumptions or beliefs

No Comments

No Known Activations