INDEX

Explanations

words and phrases indicating causality or condition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

 hatta

-0.17

 whom

-0.14

eux

-0.14

Î¯Î³

-0.14

/or

-0.14

ovah

-0.14

quete

-0.13

 [âĢ¦

-0.13

 herself

-0.13

.btnExit

-0.13

POSITIVE LOGITS

0.23

 there

0.22

forth

0.19

there

0.18

lest

0.16

 they

0.16

,↵

0.16

 ÙģØ¥ÙĨ

0.16

it

0.16

,it

0.16

Activations Density 0.268%

words and phrases indicating causality or condition

No Comments

No Known Activations