INDEX

Explanations

phrases indicating lack of options or being forced into a particular action

oai_token-act-pair · gpt-3.5-turbo

phrases related to having limited options or decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

urdy

-0.63

 Roose

-0.63

yip

-0.62

govtrack

-0.60

Interstitial

-0.59

alus

-0.59

 aggregation

-0.59

ividual

-0.59

ippi

-0.57

orf

-0.57

POSITIVE LOGITS

 whatsoever

0.90

BIL

0.78

but

0.76

ices

0.76

DO

0.72

 except

0.71

 anymore

0.69

 âĸº

0.67

llor

0.66

enegger

0.66

Activations Density 0.073%

phrases indicating lack of options or being forced into a particular action

phrases related to having limited options or decisions

No Comments

No Known Activations