INDEX

Explanations

references to individuals or groups, particularly focusing on "who" 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 @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.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

oksen

-0.17

itez

-0.14

zte

-0.14

ÑħÑĥ

-0.13

ÑĢÐ°Ð²Ð°

-0.13

gba

-0.13

ston

-0.13

arkan

-0.13

&eacute

-0.13

ÑģÑĤÐµÑĢ

-0.13

POSITIVE LOGITS

 otherwise

0.21

otherwise

0.18

Otherwise

0.18

 Otherwise

0.17

osh

0.16

 OTHERWISE

0.15

 might

0.15

agar

0.15

Activations Density 0.158%

references to individuals or groups, particularly focusing on "who" in various contexts

No Comments

No Known Activations