INDEX

Explanations

instances of the word "interfere" and its variations, indicating a focus on disruptions or interruptions in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

gger

-0.16

Ð½

-0.15

ÑĤÐ¾ÑĢ

-0.15

setter

-0.15

igned

-0.14

rame

-0.14

 latter

-0.14

agar

-0.14

ssi

-0.14

POSITIVE LOGITS

å¼ı

0.18

prise

0.16

ative

0.16

ives

0.15

Ãªncia

0.15

_sdk

0.15

indre

0.15

 Rhodes

0.14

 between

0.14

hyth

0.14

Activations Density 0.065%

instances of the word "interfere" and its variations, indicating a focus on disruptions or interruptions in various contexts

No Comments

No Known Activations