INDEX

Explanations

escape sequences and regular expression patterns

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.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

lique

-0.16

OLON

-0.15

quote

-0.15

é¼»

-0.15

ilton

-0.14

rint

-0.14

fect

-0.14

ettle

-0.14

 IonicModule

-0.14

thon

-0.14

POSITIVE LOGITS

0.19

0.16

å¼ı

0.15

izzo

0.15

arov

0.15

 fair

0.15

unately

0.14

UMB

0.14

Activations Density 0.012%

escape sequences and regular expression patterns

No Comments

No Known Activations