INDEX

Explanations

phrases related to shedding light or revealing information

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "shed" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

anwhile

-0.73

Pav

-0.70

azel

-0.64

odore

-0.61

ierra

-0.58

 protected

-0.57

War

-0.57

ONES

-0.57

 Fifth

-0.57

 Orion

-0.55

POSITIVE LOGITS

 shed

1.34

 sheds

1.10

 shedding

1.06

 Shed

0.89

ittal

0.88

shed

0.85

brid

0.83

iencies

0.76

yards

0.75

ouse

0.74

Activations Density 0.009%

phrases related to shedding light or revealing information

mentions of the word "shed" in various contexts

No Comments

No Known Activations