INDEX

Explanations

words related to discussion or commentary on various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

xing

-0.16

aket

-0.16

ÑıÑĩ

-0.14

ible

-0.14

áº¡

-0.13

icates

-0.13

Hou

-0.13

ahkan

-0.13

icky

-0.13

rone

-0.13

POSITIVE LOGITS

/from

0.16

_unused

0.15

tml

0.15

Ð²Ðµ

0.15

fur

0.14

anos

0.14

İĺìĿ´

0.14

wie

0.14

sb

0.14

((!

0.13

Activations Density 0.141%

words related to discussion or commentary on various topics

No Comments

No Known Activations