INDEX

Explanations

multiple uses of the word "more" and related intensifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

ponsible

-0.16

ardo

-0.16

hani

-0.15

icros

-0.15

Ð½Ð¸Ð²

-0.15

animate

-0.15

udad

-0.14

rome

-0.14

idious

-0.14

erten

-0.14

POSITIVE LOGITS

 completely

0.26

 totally

0.21

 completamente

0.21

å®Įåħ¨

0.20

 fully

0.20

 Completely

0.19

 often

0.18

-effect

0.18

fully

0.18

 surely

0.18

Activations Density 0.073%

multiple uses of the word "more" and related intensifiers

No Comments

No Known Activations