INDEX

Explanations

pronouns followed by a verb or finite verb phrase

oai_token-act-pair · gpt-3.5-turbo

phrases or statements emphasizing strong assertions or conclusions

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

ARS

-0.78

aws

-0.74

amia

-0.74

pots

-0.72

 Streets

-0.72

aths

-0.71

iors

-0.71

icons

-0.71

tones

-0.71

anches

-0.70

POSITIVE LOGITS

 phenomenon

1.20

 discrepancy

1.10

 latter

1.03

 arrangement

1.00

 article

0.96

 newfound

0.95

 trope

0.95

 assertion

0.94

 complication

0.91

 particular

0.91

Activations Density 0.173%

pronouns followed by a verb or finite verb phrase

phrases or statements emphasizing strong assertions or conclusions

No Comments

No Known Activations