INDEX

Explanations

references to reality television and personal sentiments

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

okud

-0.17

eza

-0.16

eday

-0.14

 çģ

-0.14

Ãłm

-0.14

assin

-0.14

conto

-0.14

ivent

-0.14

edis

-0.14

kova

-0.14

POSITIVE LOGITS

 Channel

0.18

 Season

0.17

 reality

0.17

@protocol

0.16

 channel

0.16

Channel

0.15

 cameras

0.15

 show

0.15

 Reality

0.15

çĢ¬

0.15

Activations Density 0.108%

references to reality television and personal sentiments

No Comments

No Known Activations