INDEX

Explanations

specific version numbers or software versions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

rganization

-0.17

 DevComponents

-0.16

esan

-0.15

acula

-0.15

èĻ

-0.15

allery

-0.14

ilter

-0.14

attery

-0.14

inox

-0.14

erra

-0.14

POSITIVE LOGITS

 Crowley

0.15

 Seah

0.15

GHz

0.14

ìłĪ

0.14

Wal

0.14

achel

0.14

 Transitional

0.14

RC

0.14

 mitig

0.13

GHz

0.13

Activations Density 0.039%

specific version numbers or software versions

No Comments

No Known Activations