INDEX

Explanations

phrases and descriptions related to intuitive features or solutions

oai_token-act-pair · gpt-3.5-turbo

terms related to intuition and intuitive understanding

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Companies

-0.71

sg

-0.66

gre

-0.65

thin

-0.65

abad

-0.64

metal

-0.64

mington

-0.64

emetery

-0.64

ppo

-0.63

onies

-0.63

POSITIVE LOGITS

 intuitive

1.26

uitive

1.20

ly

1.06

istically

0.96

 disson

0.90

 intuition

0.89

 minded

0.88

intuitive

0.88

 instinct

0.85

ically

0.84

Activations Density 0.007%

phrases and descriptions related to intuitive features or solutions

terms related to intuition and intuitive understanding

No Comments

No Known Activations