INDEX

Explanations

references to political or governmental figures and events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

osyal

-0.15

uxt

-0.14

aan

-0.14

ypad

-0.14

ordan

-0.13

 iParam

-0.13

jee

-0.13

ÙĤØ§Ùħ

-0.13

orsch

-0.13

 Partition

-0.13

POSITIVE LOGITS

FG

0.18

 Presidency

0.17

EF

0.17

 presidency

0.17

 Rivers

0.16

 court

0.16

 President

0.16

EF

0.16

 Tamb

0.16

ï»¿#

0.15

Activations Density 0.033%

references to political or governmental figures and events

No Comments

No Known Activations