INDEX

Explanations

expressions of apology and regret

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 17-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L17R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.17.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

MMdd

-0.17

Î¸Î±

-0.15

odge

-0.15

etwork

-0.15

eko

-0.15

íķĳ

-0.14

MV

-0.14

orna

-0.14

GenerationStrategy

-0.14

 Observatory

-0.13

POSITIVE LOGITS

 Ideal

0.17

 Lesson

0.15

Reg

0.15

 Pain

0.15

 ideal

0.15

tae

0.14

olin

0.14

 overs

0.14

Vul

0.14

uxt

0.14

Activations Density 0.046%

expressions of apology and regret

No Comments

No Known Activations