INDEX

Explanations

repeated phrases or connections in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.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

omi

-0.17

oulos

-0.15

ule

-0.15

ULE

-0.15

rows

-0.14

in

-0.14

abwe

-0.14

 altern

-0.14

ahr

-0.14

POSITIVE LOGITS

 aspect

0.18

chez

0.16

 approach

0.16

 piece

0.16

lub

0.15

radu

0.15

tid

0.14

ãĥĭãĤ¢

0.14

ched

0.14

CrLf

0.14

Activations Density 0.139%

repeated phrases or connections in context

No Comments

No Known Activations