INDEX

Explanations

repeated phrases and references to past actions or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

ema

-0.17

avig

-0.15

ãĤħ

-0.15

Ã¡lo

-0.15

opro

-0.14

 Little

-0.14

 Computational

-0.14

ISTER

-0.14

isay

-0.14

dy

-0.14

POSITIVE LOGITS

Ã©li

0.17

Ð¾Ð¸ÑĤ

0.15

èª

0.15

ourke

0.14

Ð¾ÑģÑĢÐµÐ´

0.14

zech

0.14

Ð¾ÑĪ

0.14

suy

0.14

nger

0.13

umbed

0.13

Activations Density 0.005%

repeated phrases and references to past actions or concepts

No Comments

No Known Activations