INDEX

Explanations

proper nouns and names

oai_token-act-pair · gpt-3.5-turbo

specific prominent names or entities in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

owship

-0.72

LOC

-0.56

ailability

-0.55

ulner

-0.54

ominium

-0.53

terness

-0.52

forcement

-0.51

icit

-0.50

uilding

-0.50

catentry

-0.50

POSITIVE LOGITS

?),

0.92

?)

0.78

?).

0.74

 coincidence

0.72

brainer

0.68

#$

0.65

 understatement

0.64

 hadn

0.64

!?

0.64

 typo

0.63

Activations Density 1.323%

proper nouns and names

specific prominent names or entities in various contexts

No Comments

No Known Activations