INDEX

Explanations

references to trends or patterns of behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

bon

-0.16

Legacy

-0.15

apult

-0.15

ypse

-0.15

bons

-0.15

mq

-0.15

quake

-0.14

 Levin

-0.14

ute

-0.14

strcasecmp

-0.14

POSITIVE LOGITS

ential

0.15

ervers

0.15

eview

0.15

ostat

0.14

 ÎłÎ±Î½

0.14

gle

0.14

 hazards

0.14

bib

0.14

.cent

0.14

SI

0.14

Activations Density 0.005%

references to trends or patterns of behavior

No Comments

No Known Activations