INDEX

Explanations

keywords related to taking action or making a stance

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing taking a stand or support for a cause

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

ricular

-0.86

nesota

-0.72

olesc

-0.70

iliary

-0.70

EStreamFrame

-0.69

vertis

-0.69

illian

-0.68

ffic

-0.66

abyte

-0.65

alez

-0.65

POSITIVE LOGITS

still

0.99

 stand

0.98

Stand

0.83

ings

0.75

pants

0.74

BuyableInstoreAndOnline

0.74

 tall

0.73

Standing

0.72

 onstage

0.71

hof

0.71

Activations Density 0.017%

keywords related to taking action or making a stance

phrases emphasizing taking a stand or support for a cause

No Comments

No Known Activations