INDEX

Explanations

instances of the word "rep" in various forms, indicating repetition or redoing actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

ertoire

-0.19

onis

-0.18

roll

-0.17

eu

-0.17

ea

-0.17

eb

-0.17

kowski

-0.16

ernet

-0.16

Ø±

-0.16

Interop

-0.16

POSITIVE LOGITS

aid

0.21

op

0.20

atri

0.20

ack

0.20

sons

0.19

orp

0.19

ud

0.18

ivot

0.17

ug

0.17

rior

0.17

Activations Density 0.009%

instances of the word "rep" in various forms, indicating repetition or redoing actions

No Comments

No Known Activations