INDEX

Explanations

names related to grants

oai_token-act-pair · gpt-3.5-turbo

references to the word "Grant" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

thel

-0.66

 abdom

-0.65

 pneum

-0.63

acters

-0.62

 Beaut

-0.61

 ammonia

-0.61

 Daylight

-0.60

 fats

-0.59

otta

-0.58

Ïī

-0.57

POSITIVE LOGITS

ees

1.25

ee

1.09

 Thornton

0.90

aire

0.88

lee

0.85

zeb

0.82

elson

0.78

ibal

0.76

ingham

0.73

 recipients

0.73

Activations Density 0.042%

names related to grants

references to the word "Grant" and related terms

No Comments

No Known Activations