INDEX

Explanations

requests for users to obtain more information or explore additional content

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 22-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L22R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.22.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

essler

-0.16

cec

-0.15

èģŀ

-0.14

Ð¾ÑĪ

-0.14

Overlay

-0.14

elden

-0.14

ÑĥÑĪ

-0.14

 Reddit

-0.13

gens

-0.13

 reddit

-0.13

POSITIVE LOGITS

 pages

0.20

 below

0.19

 navigation

0.17

our

0.17

ă

0.17

below

0.16

pages

0.15

Below

0.15

our

0.14

 navigating

0.14

Activations Density 0.137%

requests for users to obtain more information or explore additional content

No Known Activations