INDEX

Explanations

evaluative statements regarding the quality or significance of various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

aptcha

-0.15

ixa

-0.15

erce

-0.15

éĻµ

-0.15

ohen

-0.15

igue

-0.14

bedo

-0.14

ÙģØ§Ø¹

-0.14

/goto

-0.14

uku

-0.14

POSITIVE LOGITS

iw

0.18

aday

0.17

elik

0.16

osc

0.16

é¼»

0.15

SKI

0.15

oz

0.15

 Briggs

0.15

Mor

0.15

Activations Density 0.168%

evaluative statements regarding the quality or significance of various subjects

No Comments

No Known Activations