INDEX

Explanations

quantitative data or numerical values in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

 Tune

-0.20

etting

-0.18

rosso

-0.15

et

-0.14

iya

-0.14

flammatory

-0.14

ofile

-0.14

yne

-0.14

CLUDING

-0.14

 lÄ©nh

-0.13

POSITIVE LOGITS

ĨĴ

0.17

ces

0.16

wand

0.15

uct

0.15

witter

0.15

veau

0.14

uter

0.14

ê·¹

0.14

rÃ¥

0.14

ama

0.14

Activations Density 0.005%

quantitative data or numerical values in context

No Comments

No Known Activations

quantitative data or numerical values in context

No Comments

No Known Activations