INDEX

Explanations

symbols or indicators related to coding or data structure formatting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

erin

-0.16

lou

-0.14

 Individual

-0.14

 Rockefeller

-0.14

é§

-0.14

jon

-0.13

 Bras

-0.13

olet

-0.13

bury

-0.13

orn

-0.13

POSITIVE LOGITS

åĪĩ

0.15

ervers

0.14

 Template

0.14

uids

0.14

}-

0.14

odium

0.14

LineColor

0.14

estroy

0.14

_HEX

0.14

âĸ²

0.14

Activations Density 0.001%

symbols or indicators related to coding or data structure formatting

No Comments

No Known Activations