INDEX

Explanations

places, locations, and geographical information

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to geographic locations and their conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

oxide

-0.81

payers

-0.78

ocide

-0.78

Ç

-0.77

orial

-0.77

ality

-0.76

employ

-0.75

APH

-0.75

payer

-0.74

anova

-0.74

POSITIVE LOGITS

 vicinity

1.30

 valleys

1.02

 midst

1.01

 mountains

0.98

 evenings

0.98

 footh

0.96

 canyon

0.95

 woods

0.94

 distance

0.93

 hills

0.93

Activations Density 0.169%

places, locations, and geographical information

references to geographic locations and their conditions

No Comments

No Known Activations