INDEX

Explanations

references to academic subjects or disciplines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 19-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L19R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.19.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

oman

-0.17

ampus

-0.16

 Claw

-0.15

IGHL

-0.15

assel

-0.15

ikon

-0.15

ullo

-0.14

 ëĪĦêµ¬

-0.14

Xt

-0.14

 Hlav

-0.14

POSITIVE LOGITS

Insensitive

0.16

quine

0.15

 Kenny

0.15

å¹¸

0.15

0.14

ois

0.14

/full

0.14

eness

0.14

 coarse

0.14

PEC

0.14

Activations Density 0.002%

references to academic subjects or disciplines

No Comments

No Known Activations