INDEX

Explanations

strong verbs and action-related words 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 @ 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

ahn

-0.17

atch

-0.14

egal

-0.14

aran

-0.14

emoc

-0.14

Lon

-0.13

 postpone

-0.13

urai

-0.13

anders

-0.13

 dagen

-0.13

POSITIVE LOGITS

etc

0.27

 ÑĤÐ¾ÑīÐ¾

0.21

etc

0.18

illac

0.15

dit

0.15

áº

0.15

ffects

0.14

eson

0.14

agnosis

0.14

ä»¥åıĬ

0.14

Activations Density 0.193%

strong verbs and action-related words in various contexts

No Comments

No Known Activations