INDEX

Explanations

phrases related to solving puzzles or mysteries

oai_token-act-pair · gpt-3.5-turbo

terms related to unraveling complex problems or mysteries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

eworld

-0.69

rera

-0.68

amic

-0.63

 Titus

-0.62

lav

-0.61

alla

-0.61

Downloadha

-0.61

cca

-0.60

cer

-0.60

gar

-0.60

POSITIVE LOGITS

 unravel

1.20

edIn

1.00

ĸļ

0.87

ing

0.86

ingly

0.78

 Leilan

0.73

ed

0.70

angling

0.70

lement

0.70

 decipher

0.70

Activations Density 0.011%

phrases related to solving puzzles or mysteries

terms related to unraveling complex problems or mysteries

No Comments

No Known Activations