INDEX

Explanations

names of individuals and proper nouns

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

zÅĳ

-0.15

STALL

-0.14

Ð¾ÑĤÐµ

-0.14

Ð½Ð°ÑĢ

-0.14

contri

-0.14

rightness

-0.13

vrd

-0.13

mai

-0.13

stalk

-0.13

à¸²à¸ł

-0.13

POSITIVE LOGITS

Lag

0.17

oby

0.15

ensored

0.14

/or

0.14

Cul

0.14

 alike

0.14

lag

0.14

sl

0.14

 Loft

0.14

lag

0.13

Activations Density 0.142%

names of individuals and proper nouns

No Comments

No Known Activations