INDEX

Explanations

references to the word "who" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

mente

-0.23

ted

-0.19

ting

-0.16

Ð½ÐµÐ½

-0.15

ning

-0.15

ned

-0.15

gado

-0.15

IEL

-0.15

uen

-0.14

erais

-0.14

POSITIVE LOGITS

 else

0.24

_else

0.16

oping

0.16

soever

0.16

osh

0.16

TestFixture

0.15

apk

0.14

oles

0.14

	else

0.14

beer

0.14

Activations Density 0.033%

references to the word "who" in various contexts

No Comments

No Known Activations