INDEX

Explanations

dominant and frequently occurring words and phrases related to connections and relationships within context

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

itch

-0.17

erb

-0.17

uye

-0.15

Ð¸ÑĩÐ°

-0.15

aha

-0.15

olulu

-0.14

íĥ

-0.14

ocio

-0.14

/projects

-0.14

ì¤Ģ

-0.14

POSITIVE LOGITS

alama

0.16

à¤Łà¤ķ

0.15

sel

0.15

Å¡k

0.14

 Cannon

0.14

ModelProperty

0.14

ictor

0.14

encer

0.14

nÃ½

0.13

 Tester

0.13

Activations Density 0.001%

dominant and frequently occurring words and phrases related to connections and relationships within context

No Comments

No Known Activations