INDEX

Explanations

repetitive or impactful phrases and concepts related to distribution and limitations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

anganese

-0.17

posables

-0.15

oin

-0.15

FieldValue

-0.15

\GeneratedValue

-0.15

plex

-0.14

Dup

-0.14

å¸Ī

-0.14

jom

-0.14

geh

-0.14

POSITIVE LOGITS

eck

0.19

_blocking

0.15

áº½

0.15

 Buen

0.15

 Factor

0.14

DAT

0.14

loh

0.14

åł

0.14

Gri

0.14

ceptions

0.14

Activations Density 0.021%

repetitive or impactful phrases and concepts related to distribution and limitations

No Comments

No Known Activations