INDEX

Explanations

variations of the word "absorb" and related terms

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

oog

-0.18

aurus

-0.18

ying

-0.16

amura

-0.16

announce

-0.15

throat

-0.15

Ã¼Ã§Ã¼k

-0.15

erializer

-0.15

çĲĨ

-0.15

lek

-0.15

POSITIVE LOGITS

ence

0.31

urd

0.30

olut

0.30

OLUTE

0.29

Abs

0.29

abs

0.27

orption

0.27

ences

0.27

orb

0.27

cess

0.26

Activations Density 0.009%

variations of the word "absorb" and related terms

No Comments

No Known Activations