INDEX

Explanations

No Explanations Found

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

awan

-0.15

anz

-0.15

ect

-0.15

icker

-0.15

 ç¢

-0.14

reu

-0.14

Ì£

-0.14

 subsequently

-0.14

ickers

-0.14

 bang

-0.13

POSITIVE LOGITS

owitz

0.14

abor

0.14

ften

0.14

 unary

0.14

IENCE

0.13

.AutoSizeMode

0.13

holm

0.13

azio

0.13

 hala

0.13

.cloudflare

0.13

Activations Density 0.012%

No Comments

No Known Activations