INDEX

Explanations

words or phrases that convey richness or complexity

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

ê³Ħ

-0.15

oren

-0.14

 Popular

-0.14

 ê³Ħ

-0.14

 popular

-0.14

 Dimit

-0.14

 ÄĳÃ¬nh

-0.14

 Poly

-0.13

mid

-0.13

My

-0.13

POSITIVE LOGITS

ults

0.16

ovÃŃ

0.15

ektor

0.15

Äįas

0.15

_Cmd

0.14

inz

0.14

 spis

0.14

laden

0.14

omain

0.14

}\.[

0.14

Activations Density 0.005%

words or phrases that convey richness or complexity

No Comments

No Known Activations