INDEX

Explanations

instances of the word "telegram."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the token "te" in various forms

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

etheless

-0.73

CVE

-0.71

DragonMagazine

-0.71

 Corruption

-0.68

å§«

-0.66

 Hawks

-0.66

CVE

-0.63

 Timberwolves

-0.63

 Orbit

-0.63

 demographics

-0.62

POSITIVE LOGITS

legram

1.35

legraph

1.24

pee

1.11

achable

1.08

quila

1.04

brate

0.98

asers

0.97

eming

0.95

uro

0.93

cn

0.91

Activations Density 0.011%

instances of the word "telegram."

occurrences of the token "te" in various forms

No Comments

No Known Activations