INDEX

Explanations

conjunctions, particularly "and," which suggests a focus on connections or cumulative contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

abcdefgh

-0.15

.gridColumn

-0.15

lope

-0.14

Ð¾Ð³Ð¾Ð´

-0.14

.synthetic

-0.14

 Pasadena

-0.14

Ð¼Ð¼

-0.14

åĺī

-0.14

Subsystem

-0.14

rocessing

-0.14

POSITIVE LOGITS

etc

0.19

Kok

0.17

/or

0.16

APS

0.16

rens

0.15

omik

0.15

 Mothers

0.15

etc

0.14

 overall

0.14

jar

0.14

Activations Density 0.169%

conjunctions, particularly "and," which suggests a focus on connections or cumulative contexts

No Comments

No Known Activations