INDEX

Explanations

phrases indicating a cause-and-effect relationship or dependency on certain conditions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating conditionality or dependency

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

vision

-0.82

spir

-0.76

anas

-0.73

oop

-0.71

iverpool

-0.71

jc

-0.70

tha

-0.70

ãĥīãĥ©

-0.70

OWN

-0.69

shr

-0.69

POSITIVE LOGITS

 upon

0.79

awaru

0.78

 critically

0.78

 heavily

0.78

ymm

0.75

 solely

0.72

 adversely

0.69

 principally

0.69

 depend

0.68

 chiefly

0.68

Activations Density 0.016%

phrases indicating a cause-and-effect relationship or dependency on certain conditions

phrases indicating conditionality or dependency

No Comments

No Known Activations