INDEX

Explanations

phrases indicating comparison or relative degree

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

ivet

-0.17

ä½ľä¸º

-0.16

ivot

-0.16

 FileAccess

-0.14

itele

-0.14

anic

-0.14

_mC

-0.14

.Interop

-0.14

otle

-0.14

agle

-0.14

POSITIVE LOGITS

 nearly

0.23

 Nearly

0.21

Nearly

0.21

 tÃ£o

0.19

bit

0.17

pez

0.16

 Wass

0.16

rena

0.15

Has

0.15

asm

0.15

Activations Density 0.017%

phrases indicating comparison or relative degree

No Comments

No Known Activations