INDEX

Explanations

titles and headings of articles or sections

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

Ã©gor

-0.17

azor

-0.14

ÃĹ↵↵

-0.14

umbo

-0.14

fÃ¼g

-0.14

etri

-0.13

 kapÄ±

-0.13

 ná»¯a

-0.13

utto

-0.13

Ã¢n

-0.13

POSITIVE LOGITS

DEF

0.15

.xhtml

0.14

igkeit

0.14

inel

0.14

 Which

0.13

{{{

0.13

_:

0.13

_<

0.13

º

0.13

Activations Density 0.072%

titles and headings of articles or sections

No Comments

No Known Activations

titles and headings of articles or sections

No Comments

No Known Activations