INDEX

Explanations

words related to additional or extra information and details

oai_token-act-pair · gpt-3.5-turbo

the word "Additionally" and related phrases that introduce extra information

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.73

 hearts

-0.72

ger

-0.69

venge

-0.68

 Balkans

-0.65

girls

-0.64

fell

-0.63

liest

-0.63

 Fare

-0.62

ender

-0.61

POSITIVE LOGITS

æ©Ł

0.76

 noteworthy

0.69

 importantly

0.68

orthy

0.68

 guiActiveUn

0.66

aeda

0.65

notation

0.63

 referen

0.63

 utilized

0.61

poons

0.60

Activations Density 0.021%

words related to additional or extra information and details

the word "Additionally" and related phrases that introduce extra information

No Comments

No Known Activations