INDEX

Explanations

proper nouns related to hotels or other establishments

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

the word "Cas" 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 @ 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

 mishand

-0.60

 stakes

-0.58

 circulating

-0.56

 sentient

-0.56

 victorious

-0.55

 ancest

-0.53

 judging

-0.52

 ultrasound

-0.52

 fielding

-0.52

 unse

-0.52

POSITIVE LOGITS

anova

1.41

imir

1.38

hews

1.27

abl

1.25

illas

1.23

sels

1.20

hew

1.20

bah

1.19

inos

1.13

ually

1.09

Activations Density 0.030%

proper nouns related to hotels or other establishments

the word "Cas" in various contexts

No Comments

No Known Activations