INDEX

Explanations

instances of the word "been" in different contexts, indicating ongoing actions or states

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

 Inherits

-0.16

erton

-0.16

ÏĦÎŃ

-0.15

å¨ľ

-0.15

 Mour

-0.14

akhir

-0.14

apyrus

-0.14

 talented

-0.14

 Steps

-0.13

 beginnings

-0.13

POSITIVE LOGITS

 around

0.20

 Around

0.19

Around

0.18

around

0.17

 fixtures

0.15

 active

0.15

 providing

0.15

 serving

0.15

hon

0.14

 fixture

0.14

Activations Density 0.055%

instances of the word "been" in different contexts, indicating ongoing actions or states

No Comments

No Known Activations