INDEX

Explanations

references to food and meal-related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ors

-0.21

leys

-0.16

ians

-0.15

ters

-0.15

cker

-0.15

seo

-0.15

nen

-0.14

isko

-0.14

ingly

-0.14

iana

-0.14

POSITIVE LOGITS

/sn

0.20

stuff

0.17

Porn

0.16

ruba

0.16

PAD

0.15

/movie

0.15

kil

0.15

ãģ¹ãģ¦

0.15

/feed

0.15

 Porn

0.15

Activations Density 0.050%

references to food and meal-related concepts

No Comments

No Known Activations