INDEX

Explanations

phrases mentioning "ETS"

oai_token-act-pair · gpt-3.5-turbo

references to various forms of "sets" or categories

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 livest

-0.85

 charact

-0.76

 streng

-0.76

 unden

-0.73

 Witcher

-0.73

ãĥ¼ãĥĨ

-0.70

 exha

-0.70

 princ

-0.69

 suspic

-0.67

 gobl

-0.66

POSITIVE LOGITS

wana

1.19

cape

1.04

pace

0.99

onic

0.97

tered

0.93

chen

0.93

tering

0.90

TING

0.90

heet

0.90

tle

0.89

Activations Density 0.035%

phrases mentioning "ETS"

references to various forms of "sets" or categories

No Comments

No Known Activations