INDEX

Explanations

phrases indicating repetition or consistency across time frames

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 5-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L5R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.5.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

lege

-0.19

 Tess

-0.18

 Mandela

-0.16

hips

-0.14

lein

-0.14

gro

-0.14

Tah

-0.14

ossier

-0.14

ko

-0.13

oh

-0.13

POSITIVE LOGITS

opal

0.16

.scalablytyped

0.15

toJson

0.15

emouth

0.15

ardy

0.14

ayne

0.14

Isl

0.14

.bam

0.13

terra

0.13

ottom

0.13

Activations Density 0.037%

phrases indicating repetition or consistency across time frames

No Comments

No Known Activations