INDEX

Explanations

terms related to eating disorders

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ore."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

srf

-0.80

ulk

-0.73

cffff

-0.73

otaur

-0.70

¥ŀ

-0.68

ilts

-0.68

arb

-0.68

itars

-0.68

insula

-0.68

ued

-0.66

POSITIVE LOGITS

tto

1.29

gon

1.10

tsky

1.08

byss

1.05

lli

1.05

ttes

1.04

xia

0.96

nz

0.95

tta

0.95

cki

0.90

Activations Density 0.020%

terms related to eating disorders

instances of the word "ore."

No Comments

No Known Activations