INDEX

Explanations

locations or cities, particularly Kiev and Moscow

oai_token-act-pair · gpt-3.5-turbo

names of cities and capitals involved in geopolitical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

lihood

-0.90

Redd

-0.80

Trend

-0.80

Merit

-0.79

series

-0.79

Torrent

-0.76

bleacher

-0.75

 Scotch

-0.75

é¾į

-0.71

Professional

-0.70

POSITIVE LOGITS

 Ankara

0.93

 embassy

0.85

 consulate

0.85

 Jinping

0.81

 coffers

0.79

hran

0.78

ervatives

0.78

 ally

0.77

iets

0.77

 bureaucracy

0.77

Activations Density 0.047%

locations or cities, particularly Kiev and Moscow

names of cities and capitals involved in geopolitical contexts

No Comments

No Known Activations