INDEX

Explanations

negative evaluations of excuses or justifications for bad behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

ropp

-0.15

ToPoint

-0.14

Ø§ÙĪØ±

-0.14

	friend

-0.13

avers

-0.13

/md

-0.13

Ã¥n

-0.13

sst

-0.13

 INTERFACE

-0.13

	explicit

-0.13

POSITIVE LOGITS

 excuse

0.93

 excuses

0.86

Exc

0.79

exc

0.75

exc

0.73

Exc

0.72

-exc

0.68

 justification

0.64

 justify

0.59

.exc

0.55

Activations Density 0.319%

negative evaluations of excuses or justifications for bad behavior

No Comments

No Known Activations