INDEX

Explanations

phrases that emphasize relationships or connections

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

agra

-0.15

nof

-0.14

allo

-0.14

ger

-0.14

ided

-0.14

SEP

-0.14

YLES

-0.13

venue

-0.13

agger

-0.13

Ø¬Ø±

-0.13

POSITIVE LOGITS

baugh

0.16

ãĥ³ãĥ

0.15

isel

0.14

ãĤ·ãĥ¥

0.14

odus

0.14

 Hind

0.14

enticator

0.14

asje

0.14

iring

0.13

oday

0.13

Activations Density 0.434%

phrases that emphasize relationships or connections

No Comments

No Known Activations