INDEX

Explanations

names and titles related to community organizations and events

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

titles and roles in organizational or community 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

 cumbers

-0.60

 confir

-0.57

bably

-0.57

 relate

-0.54

 intermitt

-0.54

 teasp

-0.53

 accustomed

-0.52

 schemes

-0.51

 hesitate

-0.51

 toler

-0.51

POSITIVE LOGITS

THE

0.88

OUR

0.78

 Excellence

0.75

AST

0.71

PHI

0.71

 UNITED

0.70

 Historic

0.69

NOR

0.69

USS

0.68

UCS

0.68

Activations Density 0.572%

names and titles related to community organizations and events

titles and roles in organizational or community contexts

No Comments

No Known Activations