INDEX

Explanations

phrases related to unexpected or significant events and changes

oai_token-act-pair · gpt-3.5-turbo

phrases involving the concept of "turns of events."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

imentary

-0.74

VICE

-0.69

raq

-0.65

ļéĨĴ

-0.58

ded

-0.58

Que

-0.57

RAG

-0.56

der

-0.56

umerable

-0.55

bats

-0.55

POSITIVE LOGITS

phrase

0.74

 pace

0.72

 administrations

0.65

 scenery

0.64

EVA

0.64

anship

0.64

 phrase

0.64

Listener

0.63

uncture

0.61

 tempo

0.60

Activations Density 0.105%

phrases related to unexpected or significant events and changes

phrases involving the concept of "turns of events."

No Comments

No Known Activations