INDEX

Explanations

the word "earlier"

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

thalm

-0.15

buster

-0.15

onga

-0.15

odom

-0.15

arius

-0.14

Ã²ng

-0.14

ãģĵãģĿ

-0.14

rxjs

-0.14

kraine

-0.14

imir

-0.13

POSITIVE LOGITS

andi

0.15

ilyn

0.15

 Tamb

0.15

oro

0.14

ay

0.14

ures

0.14

ABCDEFG

0.14

ayla

0.14

enn

0.14

ensa

0.13

Activations Density 0.024%

the word "earlier"

No Comments

No Known Activations