INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

fsp

-0.15

_expire

-0.14

rego

-0.14

åĽ

-0.13

_super

-0.13

 ÙģØ¨Ø±Ø§ÙĬØ±

-0.13

à¹Ģà¸ªà¸£

-0.13

ayan

-0.13

animations

-0.13

hart

-0.12

POSITIVE LOGITS

ijkstra

0.18

enton

0.17

lec

0.16

 situation

0.15

eshire

0.15

Ø§ÙģÙĩ

0.14

sit

0.14

orde

0.14

/browse

0.14

éĹ

0.14

Activations Density 0.000%

No Known Activations

This feature has no known activations.