INDEX

Explanations

phrases expressing uncertainty or needing to wait for further information

oai_token-act-pair · gpt-3.5-turbo

highlighted future actions or predictions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

efined

-0.69

SPONSORED

-0.65

emin

-0.64

ensable

-0.62

Description

-0.62

Newsletter

-0.61

Introduced

-0.60

places

-0.60

rawdownloadcloneembedreportprint

-0.60

DERR

-0.58

POSITIVE LOGITS

 wait

1.20

 hope

0.98

 await

0.98

 hoping

0.94

 speculate

0.94

 rejoice

0.88

 beware

0.87

 dunno

0.87

 guess

0.84

 WATCH

0.84

Activations Density 0.327%

phrases expressing uncertainty or needing to wait for further information

highlighted future actions or predictions

No Comments

No Known Activations