INDEX

Explanations

phrases related to policy proposals or possible actions

oai_token-act-pair · gpt-3.5-turbo

keywords related to legislation and policy changes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

©¶æ¥µ

-0.70

Romney

-0.62

Pope

-0.59

Els

-0.59

Lenin

-0.58

Enlarge

-0.58

Ocean

-0.57

Mods

-0.56

Wan

-0.56

affer

-0.56

POSITIVE LOGITS

 this

0.97

the

0.97

 either

0.79

 both

0.77

any

0.71

our

0.70

 these

0.70

 behalf

0.70

 your

0.65

 what

0.64

Activations Density 0.685%

phrases related to policy proposals or possible actions

keywords related to legislation and policy changes

No Comments

No Known Activations