INDEX

Explanations

terms related to physical or conceptual divisions or separations

oai_token-act-pair · gpt-3.5-turbo

terms related to political or social divisions

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

inel

-0.81

kinson

-0.76

anamo

-0.75

psc

-0.74

odium

-0.73

enez

-0.68

aina

-0.67

lev

-0.66

azer

-0.66

azor

-0.65

POSITIVE LOGITS

 partition

1.23

 partitions

1.22

 divid

0.87

ãĥ¼ãĥĨ

0.83

 breakup

0.78

ipation

0.78

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.77

 flares

0.75

ãĥ¼ãĥĨãĤ£

0.75

nect

0.71

Activations Density 0.009%

terms related to physical or conceptual divisions or separations

terms related to political or social divisions

No Comments

No Known Activations