INDEX

Explanations

terms related to degeneration, deterioration, or decay

oai_token-act-pair · gpt-3.5-turbo

terms related to degrees, particularly in relation to limits or measurement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cule

-0.74

 stakes

-0.70

éĹĺ

-0.70

bley

-0.64

mberg

-0.59

 Nevada

-0.59

kers

-0.59

 crowds

-0.59

JPM

-0.59

 GOODMAN

-0.58

POSITIVE LOGITS

rees

1.35

rowth

1.28

rad

1.27

ener

1.21

gradation

0.95

deg

0.94

rave

0.89

¿

0.87

ree

0.86

reed

0.85

Activations Density 0.006%

terms related to degeneration, deterioration, or decay

terms related to degrees, particularly in relation to limits or measurement

No Comments

No Known Activations