INDEX

Explanations

locations or places

oai_token-act-pair · gpt-3.5-turbo

frequent mentions of geographical locations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

arians

-0.74

ality

-0.73

payer

-0.73

anova

-0.72

orial

-0.71

payers

-0.70

ueller

-0.69

LEASE

-0.68

STD

-0.66

inst

-0.65

POSITIVE LOGITS

 vicinity

1.48

 midst

1.37

 footh

1.25

 southwest

1.16

 southwestern

1.15

 hills

1.14

 northeast

1.14

 countryside

1.13

 northwest

1.12

 southeastern

1.12

Activations Density 0.170%

locations or places

frequent mentions of geographical locations

No Comments

No Known Activations