INDEX

Explanations

numerical indicators of rankings or performance over time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

èĸĦ

-0.16

.throw

-0.16

shal

-0.15

ãģĸ

-0.14

æ¡Ĳ

-0.14

compat

-0.14

ioni

-0.14

isel

-0.14

ificio

-0.14

 Ø±ÙĪØŃ

-0.13

POSITIVE LOGITS

 straight

0.71

 consecutive

0.60

straight

0.59

 Straight

0.57

 consec

0.52

Straight

0.52

 consect

0.47

 successive

0.47

è¿ŀ

0.44

-con

0.44

Activations Density 0.109%

numerical indicators of rankings or performance over time

No Comments

No Known Activations