INDEX

Explanations

various instances and forms of the word "am" and numerical quantities, including measurements and counts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

Ð»ÑĥÐ³Ð¸

-0.15

itemap

-0.15

agara

-0.15

Ã¤ter

-0.14

	Copyright

-0.14

ovich

-0.14

 Î¬Î»

-0.14

 Nest

-0.14

ssql

-0.13

Ã¦r

-0.13

POSITIVE LOGITS

ohan

0.16

azzi

0.16

pseudo

0.16

 ninh

0.16

illet

0.15

ãĤ¹ãĤ¯

0.15

isle

0.15

Harness

0.14

_goto

0.14

oka

0.14

Activations Density 0.004%

various instances and forms of the word "am" and numerical quantities, including measurements and counts

No Comments

No Known Activations