INDEX

Explanations

code syntax elements and function calls

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

iti

-0.15

dag

-0.15

yen

-0.14

agt

-0.14

ask

-0.14

raman

-0.14

sk

-0.14

ita

-0.14

mut

-0.13

arr

-0.13

POSITIVE LOGITS

asics

0.16

 Initiative

0.15

ÄħÅ¼

0.15

 spit

0.15

rias

0.14

uard

0.14

rist

0.13

çİ°åľº

0.13

ertia

0.13

ftar

0.13

Activations Density 0.076%

code syntax elements and function calls

No Comments

No Known Activations