INDEX

Explanations

words related to paradoxes or asymmetry

oai_token-act-pair · gpt-3.5-turbo

concepts of paradoxes and contradictions

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

azer

-0.80

awar

-0.72

oway

-0.70

GER

-0.66

ridge

-0.66

argon

-0.65

owers

-0.65

uci

-0.65

acca

-0.65

estone

-0.65

POSITIVE LOGITS

ical

0.99

ically

0.97

 between

0.94

xual

0.90

etrical

0.86

 symmetry

0.83

icals

0.82

etry

0.82

etric

0.77

yip

0.77

Activations Density 0.109%

words related to paradoxes or asymmetry

concepts of paradoxes and contradictions

No Comments

No Known Activations