INDEX

Explanations

abstract or formal phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

probably

-0.18

maybe

-0.17

 maybe

-0.16

èĢĮ

-0.16

perhaps

-0.15

hopefully

-0.15

another

-0.15

sometimes

-0.14

possibly

-0.14

legate

-0.14

POSITIVE LOGITS

 however

0.90

 However

0.76

However

0.69

 HOWEVER

0.68

however

0.64

 Ð¾Ð´Ð½Ð°ÐºÐ¾

0.60

 jedoch

0.60

 vÅ¡ak

0.49

 jednak

0.44

åį´

0.44

Activations Density 0.307%

abstract or formal phrases

No Comments

No Known Activations