INDEX

Explanations

numerical values and punctuation within the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 11-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L11R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.11.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

isse

-0.17

 blow

-0.17

nze

-0.15

anz

-0.15

iber

-0.14

Ã©mon

-0.14

Anc

-0.13

GG

-0.13

stor

-0.13

bet

-0.13

POSITIVE LOGITS

rippling

0.17

ffset

0.16

|array

0.16

MUX

0.15

ruÅ¾

0.14

ÚĨØ§Ø±

0.14

_basename

0.14

OLOR

0.14

anford

0.14

raction

0.14

Activations Density 0.003%

numerical values and punctuation within the text

No Comments

No Known Activations