INDEX

Explanations

specific character patterns or fragments often related to names or titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 10-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L10R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.10.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

ipc

-0.17

orca

-0.15

FEC

-0.15

 burge

-0.14

Ai

-0.14

jeme

-0.13

 Licensed

-0.13

iones

-0.13

 licensed

-0.13

ONUS

-0.13

POSITIVE LOGITS

-initialized

0.17

tvrt

0.16

ehler

0.16

 Tall

0.14

İ

0.14

Constructed

0.14

uard

0.14

pain

0.13

tat

0.13

Reb

0.13

Activations Density 0.009%

specific character patterns or fragments often related to names or titles

No Comments

No Known Activations