INDEX

Explanations

phrases related to being accused of something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "accused."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

emetery

-0.76

patch

-0.76

[|

-0.72

wed

-0.65

IDE

-0.62

erved

-0.62

quickShipAvailable

-0.62

ARCH

-0.62

tem

-0.62

WAR

-0.61

POSITIVE LOGITS

 accuse

0.93

 accused

0.89

 accusing

0.87

 accuses

0.83

iary

0.76

heet

0.73

isance

0.73

 suspects

0.70

ilities

0.70

 witches

0.70

Activations Density 0.015%

phrases related to being accused of something

instances of the word "accused."

No Comments

No Known Activations