INDEX

Explanations

the names and affiliations of researchers or authors in academic papers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ewart

-0.19

Å¡Å¥

-0.18

OPY

-0.15

EIF

-0.15

OMEM

-0.15

_charset

-0.14

YST

-0.14

klady

-0.14

èŃ

-0.14

POSITIVE LOGITS

 Betting

0.17

Lei

0.16

ag

0.15

enia

0.15

 stump

0.15

ctal

0.15

imp

0.15

ur

0.14

 Cent

0.14

Wei

0.13

Activations Density 0.149%

the names and affiliations of researchers or authors in academic papers

No Comments

No Known Activations