INDEX

Explanations

capitalized proper nouns, particularly names and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

 cref

-0.17

itur

-0.16

chn

-0.15

orr

-0.15

opia

-0.14

oningen

-0.14

ipl

-0.14

à¸³

-0.14

edn

-0.14

å®Ī

-0.14

POSITIVE LOGITS

Kil

0.16

auen

0.16

">//

0.15

çĬ¯

0.15

/doc

0.15

 fest

0.14

inas

0.14

Meta

0.14

lder

0.14

asti

0.13

Activations Density 0.033%

capitalized proper nouns, particularly names and organizations

No Comments

No Known Activations