INDEX

Explanations

keywords or specific phrases related to notes or annotations

oai_token-act-pair · gpt-3.5-turbo

references to "notes" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

LOS

-0.80

nom

-0.68

slaught

-0.65

wyn

-0.65

amed

-0.65

SW

-0.60

 homicide

-0.60

 multif

-0.60

eve

-0.59

naissance

-0.59

POSITIVE LOGITS

 notes

1.03

notes

1.01

 Notes

0.99

books

0.90

Notes

0.77

 note

0.75

afety

0.74

 Gained

0.72

heet

0.72

.>>

0.71

Activations Density 0.010%

keywords or specific phrases related to notes or annotations

references to "notes" in various contexts

No Comments

No Known Activations