INDEX

Explanations

proper nouns and significant names in a variety of 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

IF

-0.17

WHO

-0.16

SEE

-0.16

SEE

-0.15

zilla

-0.15

eteor

-0.15

adaÅŁ

-0.15

ewise

-0.15

gi

-0.15

wechat

-0.15

POSITIVE LOGITS

¨

0.17

 geÃ§en

0.15

etat

0.14

inge

0.14

 gáº¯n

0.14

ubb

0.14

onga

0.14

 tether

0.14

ãĤ±ãĥĥãĥĪ

0.14

anking

0.14

Activations Density 0.173%

proper nouns and significant names in a variety of contexts

No Comments

No Known Activations