INDEX

Explanations

repeated mentions of the concept of "answers."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

utch

-0.15

aims

-0.14

 convention

-0.14

inkel

-0.14

quez

-0.14

ohan

-0.14

oms

-0.14

 tÃ¼k

-0.14

ucid

-0.14

 Goth

-0.14

POSITIVE LOGITS

eru

0.15

 Crowley

0.15

ÎµÎ´

0.15

edly

0.14

 flex

0.14

flex

0.14

ãĤ¤ãĤ¹

0.14

 bÃ¬

0.14

braco

0.13

èħ¾

0.13

Activations Density 0.013%

repeated mentions of the concept of "answers."

No Comments

No Known Activations