INDEX

Explanations

punctuation and pauses in dialogue or text

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

adir

-0.16

atron

-0.15

æĢ

-0.15

tam

-0.15

anto

-0.15

hus

-0.14

.infinity

-0.14

æģĴ

-0.14

ediator

-0.14

tring

-0.14

POSITIVE LOGITS

¹

0.15

lick

0.15

oldur

0.15

ging

0.14

0.13

blick

0.13

Activations Density 0.030%

punctuation and pauses in dialogue or text

No Comments

No Known Activations