INDEX

Explanations

proper nouns and corporate names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 12-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L12R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.12.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

_OBJC

-0.14

icari

-0.14

ritt

-0.14

cliffe

-0.14

haar

-0.14

getStore

-0.14

jc

-0.13

afort

-0.13

apan

-0.13

:@""

-0.13

POSITIVE LOGITS

 Group

0.15

 Crown

0.15

AGO

0.14

Ank

0.14

 Glob

0.14

bomb

0.14

asi

0.13

 Advantage

0.13

,strlen

0.13

Activations Density 0.206%

proper nouns and corporate names

No Comments

No Known Activations