INDEX

Explanations

links or buttons to click on, typically associated with taking action or navigating through content

oai_token-act-pair · gpt-3.5-turbo

phrases focusing on clicking links or buttons

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

ãĤ¼ãĤ¦ãĤ¹

-0.85

utenberg

-0.78

ELF

-0.70

ŀ

-0.70

©¶æ¥µ

-0.70

zona

-0.68

rament

-0.66

ARB

-0.65

arnaev

-0.64

SourceFile

-0.64

POSITIVE LOGITS

 behalf

1.07

shore

0.85

cue

0.83

 autop

0.76

itiveness

0.75

coming

0.72

oice

0.70

auts

0.69

erous

0.69

 impulse

0.69

Activations Density 0.038%

links or buttons to click on, typically associated with taking action or navigating through content

phrases focusing on clicking links or buttons

No Comments

No Known Activations