INDEX

Explanations

words indicating conditions, requirements, or thresholds for achieving an effect

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.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

onte

-0.16

omain

-0.15

hift

-0.15

ÐµÑĢÐ±

-0.14

 Durant

-0.14

anja

-0.14

istency

-0.14

printw

-0.14

ektor

-0.14

KL

-0.13

POSITIVE LOGITS

ayers

0.16

ettings

0.16

overs

0.15

 Prof

0.14

oky

0.14

 irre

0.14

ento

0.13

ENTICATION

0.13

aben

0.13

ansom

0.13

Activations Density 0.004%

words indicating conditions, requirements, or thresholds for achieving an effect

No Comments

No Known Activations