INDEX

Explanations

words related to geographic locations, particularly cities or countries

oai_token-act-pair · gpt-3.5-turbo

the mention of specific geographic locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

enegger

-0.77

ãĤ¢ãĥ«

-0.76

Ö¼

-0.75

ãĥ¼ãĥ³

-0.72

å¥³

-0.69

UTION

-0.69

é¾įå¥ĳå£«

-0.67

alf

-0.66

ãĤ¼ãĤ¦ãĤ¹

-0.64

ãĥŃ

-0.64

POSITIVE LOGITS

reth

0.95

lette

0.91

vre

0.89

vernment

0.88

ng

0.87

mi

0.87

seless

0.84

hou

0.84

chet

0.83

lish

0.82

Activations Density 0.015%

words related to geographic locations, particularly cities or countries

the mention of specific geographic locations

No Comments

No Known Activations