INDEX

Explanations

words related to reassessment or re-evaluation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.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

QN

-0.15

 charged

-0.15

 context

-0.14

 McMahon

-0.14

inf

-0.14

 silence

-0.14

 Owner

-0.14

sm

-0.14

 deco

-0.13

 Boat

-0.13

POSITIVE LOGITS

inst

0.21

organization

0.19

structure

0.18

 inst

0.18

brand

0.18

bid

0.18

INST

0.17

hire

0.17

iew

0.17

IEW

0.16

Activations Density 0.028%

words related to reassessment or re-evaluation

No Comments

No Known Activations