INDEX

Explanations

references to violent attacks and terrorism

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

 Crack

-0.15

ÑħÐ¾Ð²

-0.15

 repe

-0.15

ares

-0.14

ł

-0.14

REATED

-0.14

riot

-0.13

 OVERRIDE

-0.13

xit

-0.13

ARSE

-0.13

POSITIVE LOGITS

-Ta

0.15

utomation

0.15

geist

0.15

-target

0.14

rung

0.14

ÐµÐ¼Ð¾Ð½

0.14

Ã¢m

0.14

Å¯j

0.14

 Whoever

0.14

_SKIP

0.14

Activations Density 0.079%

references to violent attacks and terrorism

No Comments

No Known Activations