INDEX

Explanations

conjunctions, particularly the word "and"

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

ertz

-0.15

 Ashe

-0.15

uros

-0.15

ä¼´

-0.14

ëĿ½

-0.14

ury

-0.14

blick

-0.14

antor

-0.14

ÄŁinden

-0.13

roit

-0.13

POSITIVE LOGITS

ì°¸

0.16

raman

0.15

 sobÄĽ

0.15

iya

0.15

 kino

0.15

ÑĢÐ°Ð³

0.14

>\<^

0.14

raquo

0.14

oggler

0.14

Activations Density 0.033%

conjunctions, particularly the word "and"

No Comments

No Known Activations