INDEX

Explanations

phrases related to decline or reduction in various metrics

oai_token-act-pair · gpt-3.5-turbo

terms related to declines or decreases in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Nev

-0.72

ZI

-0.65

POL

-0.63

ilic

-0.62

ILE

-0.62

tein

-0.61

tains

-0.61

æĿ

-0.61

Chel

-0.60

 wishes

-0.60

POSITIVE LOGITS

 sharply

0.93

 drastically

0.90

graded

0.90

 proport

0.83

 markedly

0.83

 plummet

0.81

 dramatically

0.81

 precip

0.78

ember

0.78

downs

0.76

Activations Density 0.099%

phrases related to decline or reduction in various metrics

terms related to declines or decreases in various contexts

No Comments

No Known Activations