INDEX

Explanations

references to societal labels and classifications, particularly regarding homelessness and how individuals are categorized

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

Negative Logits

/gui

-0.16

isse

-0.15

ollah

-0.15

Ð¾ÑģÐ½Ð¾Ð²

-0.15

ÐµÑģÑı

-0.14

 Bravo

-0.13

sc

-0.13

 Imagine

-0.13

Vog

-0.13

 drain

-0.13

POSITIVE LOGITS

.term

0.17

 terms

0.17

 terminology

0.17

ì¹Ń

0.16

terms

0.16

eview

0.15

Terms

0.15

 terme

0.15

CHANT

0.15

Activations Density 0.150%

references to societal labels and classifications, particularly regarding homelessness and how individuals are categorized

No Known Activations