INDEX

Explanations

names of specific entities or individuals, particularly in a news or political context

oai_token-act-pair · gpt-3.5-turbo

nouns related to institutions, organizations, and their roles 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 @ 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

asha

-0.75

awaru

-0.71

anwhile

-0.68

edIn

-0.66

annot

-0.62

 Reincarnated

-0.62

wcsstore

-0.60

 Cosponsors

-0.60

 attm

-0.59

yout

-0.59

POSITIVE LOGITS

 ample

1.10

 permission

1.08

 plenty

1.08

 pause

0.93

 considerable

0.88

an

0.88

 access

0.87

 undue

0.85

0.84

 another

0.83

Activations Density 0.152%

names of specific entities or individuals, particularly in a news or political context

nouns related to institutions, organizations, and their roles in various contexts

No Comments

No Known Activations