INDEX

Explanations

instances of the word "well" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 indeed

-0.15

reat

-0.15

å¢¨

-0.14

 juga

-0.14

ish

-0.14

 also

-0.14

ynes

-0.14

Ø§Ø¶

-0.14

IMUM

-0.14

ved

-0.14

POSITIVE LOGITS

 boot

0.66

 Boot

0.53

-boot

0.50

boot

0.49

Boot

0.48

 BOOT

0.47

_boot

0.43

/boot

0.39

 boots

0.39

BOOT

0.38

Activations Density 0.041%

instances of the word "well" in different contexts

No Comments

No Known Activations