INDEX

Explanations

descriptions or mentions of large or massive objects or entities

oai_token-act-pair · gpt-3.5-turbo

references to size and scale, particularly related to groups, entities, and quantities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

gently

-0.67

 Keller

-0.64

 Digest

-0.59

rique

-0.59

Jav

-0.58

Dak

-0.58

Cod

-0.58

Pam

-0.58

 Kelley

-0.57

 Joker

-0.57

POSITIVE LOGITS

(>

0.96

ãĤ¼ãĤ¦ãĤ¹

0.81

occup

0.80

 conglomer

0.79

unic

0.66

WER

0.66

 undertaking

0.64

————

0.63

 unden

0.62

————————

0.62

Activations Density 0.488%

descriptions or mentions of large or massive objects or entities

references to size and scale, particularly related to groups, entities, and quantities

No Comments

No Known Activations