INDEX

Explanations

comparative phrases indicating a situation or state

oai_token-act-pair · gpt-3.5-turbo

expressions of strong opinions or emotions

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

ivan

-0.65

arte

-0.64

Firstly

-0.61

iquette

-0.60

icipated

-0.60

sequently

-0.58

aeper

-0.58

received

-0.58

inguished

-0.57

osate

-0.57

POSITIVE LOGITS

 lousy

0.98

 crap

0.97

 crappy

0.96

 nasty

0.96

 weird

0.93

 dunno

0.90

 blah

0.87

 shitty

0.86

 crazy

0.86

 stuff

0.85

Activations Density 1.107%

comparative phrases indicating a situation or state

expressions of strong opinions or emotions

No Comments

No Known Activations