INDEX

Explanations

references to scientific research and publication details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

Ã²n

-0.15

ively

-0.15

eder

-0.15

-0.14

'gc

-0.14

Swe

-0.14

 sweeps

-0.14

ÐµÐµ

-0.13

 birth

-0.13

POSITIVE LOGITS

 yacc

0.17

iminal

0.16

YRO

0.15

ubb

0.14

ocoder

0.14

empo

0.14

yt

0.14

istencia

0.14

itm

0.14

ovan

0.14

Activations Density 0.005%

references to scientific research and publication details

No Comments

No Known Activations

references to scientific research and publication details

No Comments

No Known Activations