INDEX

Explanations

words related to conflict and destruction

oai_token-act-pair · gpt-3.5-turbo

specific titles or names, particularly those associated with notable entities or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ALS

-0.88

 acknow

-0.87

Pes

-0.84

 Unch

-0.83

ASC

-0.82

IB

-0.81

Um

-0.80

AH

-0.79

 Antar

-0.79

 Fitz

-0.77

POSITIVE LOGITS

organ

1.35

shop

1.34

drive

1.33

free

1.31

mere

1.29

run

1.28

bug

1.27

max

1.27

cond

1.26

floor

1.26

Activations Density 0.318%

words related to conflict and destruction

specific titles or names, particularly those associated with notable entities or events

No Comments

No Known Activations