INDEX

Explanations

references to English subtitles in various contexts

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

-0.15

ģ

-0.15

ÅĻeh

-0.15

-0.14

vais

-0.14

-slot

-0.14

stat

-0.14

ì¶ľ

-0.14

wner

-0.13

aland

-0.13

POSITIVE LOGITS

dub

0.43

 subtitle

0.43

 dubbed

0.42

 subt

0.41

dub

0.40

 captions

0.39

Dub

0.39

 caption

0.39

 subtitles

0.38

subtitle

0.37

Activations Density 0.036%

references to English subtitles in various contexts

No Known Activations