INDEX

Explanations

words related to reduction or decrease

oai_token-act-pair · gpt-3.5-turbo

terms related to reduction or decrease

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ullivan

-0.77

rote

-0.70

otle

-0.69

=-=-=-=-=-=-=-=-

-0.68

OH

-0.68

alez

-0.68

oho

-0.67

Web

-0.66

PB

-0.66

Offline

-0.65

POSITIVE LOGITS

 dimin

1.10

 diminish

0.94

 lessen

0.91

 diminished

0.80

 proport

0.80

inished

0.75

utive

0.74

Ń·

0.74

 aback

0.74

wcs

0.71

Activations Density 0.009%

words related to reduction or decrease

terms related to reduction or decrease

No Comments

No Known Activations