INDEX

Explanations

words related to starting points or beginnings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

edo

-0.14

 curs

-0.14

à¥ĩà¤¹à¤°

-0.14

bib

-0.14

Äĥ

-0.13

jin

-0.13

rms

-0.13

 cara

-0.13

 triple

-0.13

thy

-0.13

POSITIVE LOGITS

alone

0.21

hang

0.18

STRACT

0.17

ends

0.17

wick

0.15

endl

0.15

andoned

0.15

olute

0.15

rieg

0.15

LLU

0.15

Activations Density 0.006%

words related to starting points or beginnings

No Comments

No Known Activations