INDEX

Explanations

names or terms that include the letters "Ac"

oai_token-act-pair · gpt-3.5-turbo

references to a specific academic context or content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.95

EngineDebug

-0.91

è»

-0.85

SHIP

-0.85

é¾įå¥ĳå£«

-0.76

 guiActiveUnfocused

-0.75

å£«

-0.75

naire

-0.74

ĸļ

-0.74

naires

-0.73

POSITIVE LOGITS

oustic

1.36

upuncture

1.35

rylic

1.34

robat

1.13

uria

1.04

osta

1.01

acia

1.00

onso

0.92

ric

0.89

rib

0.89

Activations Density 0.014%

names or terms that include the letters "Ac"

references to a specific academic context or content

No Comments

No Known Activations