INDEX

Explanations

phrases that denote conditions, stipulations, or relationships in arguments or reasoning

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

arte

-0.18

ombs

-0.17

ieux

-0.16

arto

-0.16

iff

-0.15

ourg

-0.15

å®Ĺ

-0.15

 Rosenstein

-0.15

asted

-0.14

cest

-0.14

POSITIVE LOGITS

andi

0.16

alion

0.15

è·¡

0.14

Unnamed

0.14

instanc

0.14

.scalablytyped

0.13

ÑĢÐ°Ð·Ñĥ

0.13

 Ekon

0.13

cken

0.12

 Bench

0.12

Activations Density 0.184%

phrases that denote conditions, stipulations, or relationships in arguments or reasoning

No Comments

No Known Activations