INDEX

Explanations

mentions of the word "broadcaster"

oai_token-act-pair · gpt-3.5-turbo

mentions of different media broadcasters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

illard

-0.82

hiro

-0.69

entially

-0.66

ential

-0.64

hots

-0.64

hens

-0.64

leans

-0.63

eanor

-0.62

 Bruins

-0.61

plets

-0.61

POSITIVE LOGITS

 broadcaster

1.16

 broadcasters

0.95

casters

0.87

caster

0.85

 Broadcasting

0.83

 Broadcast

0.83

levision

0.81

 EDITION

0.78

 Seym

0.77

 broadcasting

0.75

Activations Density 0.009%

mentions of the word "broadcaster"

mentions of different media broadcasters

No Comments

No Known Activations