INDEX

Explanations

expressions of opinion or assessment about films or artistic works

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

ppers

-0.15

oine

-0.14

htdocs

-0.14

din

-0.14

apsed

-0.14

 valide

-0.14

 famously

-0.14

ilde

-0.13

®

-0.13

 notoriously

-0.13

POSITIVE LOGITS

Them

0.19

 structured

0.18

 unlike

0.18

set

0.17

 billed

0.17

 told

0.17

structured

0.17

 divided

0.17

 paced

0.17

 based

0.16

Activations Density 0.123%

expressions of opinion or assessment about films or artistic works

No Comments

No Known Activations