INDEX

Explanations

locations or places

oai_token-act-pair · gpt-3.5-turbo

references to local activities or entities

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

REF

-0.74

rer

-0.71

ger

-0.70

 Gamble

-0.68

————————

-0.67

rers

-0.67

itor

-0.66

swer

-0.63

xx

-0.62

vous

-0.62

POSITIVE LOGITS

 locally

1.02

 exting

0.98

 corrid

0.89

 localized

0.86

£ı

0.83

 brewed

0.82

 minded

0.80

 exported

0.76

 eleph

0.76

å¥³

0.76

Activations Density 0.009%

locations or places

references to local activities or entities

No Comments

No Known Activations