INDEX

Explanations

phrases or terms related to metrics or measurements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ission

-0.15

áº¯c

-0.15

enance

-0.15

 Ashton

-0.15

å¤©åłĤ

-0.15

;element

-0.15

_bh

-0.15

Ã¸d

-0.15

¼åĲĪ

-0.14

ruc

-0.14

POSITIVE LOGITS

éĸĵãģ«

0.16

ĭ

0.14

Â°

0.14

ope

0.14

ima

0.14

OLT

0.14

 Wings

0.14

ÙĪØ¨ÛĮ

0.14

é¸

0.14

Î²Î±

0.13

Activations Density 0.000%

phrases or terms related to metrics or measurements

No Comments

No Known Activations