INDEX

Explanations

terms related to fixing or repair in various contexts

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

rial

-0.17

alu

-0.16

ought

-0.15

qv

-0.15

kir

-0.14

seys

-0.14

kol

-0.14

shan

-0.14

soever

-0.14

fÃ¼

-0.14

POSITIVE LOGITS

tures

0.35

TURE

0.26

ated

0.24

er

0.22

(es

0.21

xed

0.19

gerald

0.19

ity

0.18

 broken

0.18

à¤¼

0.18

Activations Density 0.043%

terms related to fixing or repair in various contexts

No Comments

No Known Activations