INDEX

Explanations

terms related to flatness or straightness

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ikan

-0.16

holes

-0.16

imeter

-0.15

à¸°

-0.15

TECTED

-0.15

@index

-0.15

him

-0.14

itals

-0.14

sein

-0.14

à¸²à¹Ģà¸¥

-0.14

POSITIVE LOGITS

ness

0.22

-ÑĤÐ°ÐºÐ¸

0.17

ulence

0.17

ulent

0.17

ened

0.16

ities

0.16

weg

0.16

-faced

0.15

ening

0.15

bred

0.15

Activations Density 0.042%

terms related to flatness or straightness

No Comments

No Known Activations