INDEX

Explanations

specific references to TV shows and their details, such as channels, characters, and genres

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

auer

-0.14

ylene

-0.14

æ¥Ń

-0.13

ieri

-0.13

uled

-0.13

Ath

-0.13

apy

-0.13

ÑĨÑĥÐ·

-0.13

 Comm

-0.13

_X

-0.13

POSITIVE LOGITS

eyin

0.15

ÙĤØ¨

0.14

oste

0.13

pek

0.13

ÏĥÎºÎµ

0.13

Ã£n

0.13

icast

0.13

setProperty

0.13

 decis

0.13

BOOLE

0.13

Activations Density 0.093%

specific references to TV shows and their details, such as channels, characters, and genres

No Comments

No Known Activations