INDEX

Explanations

keywords related to success or impact

oai_token-act-pair · gpt-3.5-turbo

instances of the word "hit" and its variations in multiple contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

UTH

-0.70

æ©Ł

-0.66

ipation

-0.63

 Transition

-0.60

 Principles

-0.60

 Warm

-0.60

å¸

-0.59

ife

-0.59

pires

-0.59

_>

-0.58

POSITIVE LOGITS

ches

1.26

chens

1.24

ched

1.19

achi

1.11

ters

1.00

box

0.99

boxes

0.97

tle

0.87

chers

0.86

ting

0.84

Activations Density 0.030%

keywords related to success or impact

instances of the word "hit" and its variations in multiple contexts

No Comments

No Known Activations