INDEX

Explanations

references to mistaken beliefs and contradictions in arguments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

ropolis

-0.16

venes

-0.16

ses

-0.15

Ø§ÙĦØ§Øª

-0.15

flen

-0.14

ewood

-0.14

UDIO

-0.14

][/

-0.14

idor

-0.14

forder

-0.14

POSITIVE LOGITS

igm

0.16

ToOne

0.15

Ã±ana

0.14

 Martial

0.14

enti

0.14

avy

0.14

TestingModule

0.14

Ú¯ÛĮ

0.14

 pill

0.13

 heure

0.13

Activations Density 0.110%

references to mistaken beliefs and contradictions in arguments

No Comments

No Known Activations