INDEX

Explanations

elements related to architectural descriptions and structural features

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

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

itzer

-0.19

 Wire

-0.16

 wire

-0.16

 mand

-0.16

MMC

-0.15

ordova

-0.15

arch

-0.15

 Arch

-0.14

idea

-0.14

 prec

-0.14

POSITIVE LOGITS

 subs

0.20

subs

0.18

 poss

0.16

para

0.16

 existe

0.16

 appart

0.16

 figure

0.15

isco

0.15

 compt

0.15

 exist

0.15

Activations Density 0.021%

elements related to architectural descriptions and structural features

No Comments

No Known Activations