INDEX

Explanations

references to previous information and the concept of "above mentioned" items

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

ãĥ³ãĥĸ

-0.17

âĺħâĺħ

-0.14

Ã¶tt

-0.14

izin

-0.13

polator

-0.13

argent

-0.13

clair

-0.13

ois

-0.13

boru

-0.13

zac

-0.12

POSITIVE LOGITS

 mentioned

0.89

 above

0.84

 aforementioned

0.82

mentioned

0.82

above

0.73

 Above

0.63

Above

0.62

 Mention

0.62

 mention

0.59

 Ð²ÑĭÑĪÐµ

0.59

Activations Density 0.462%

references to previous information and the concept of "above mentioned" items

No Comments

No Known Activations