INDEX

Explanations

words related to comprehension and understanding, particularly in contexts where clarity or apprehension may be lacking

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.31.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

ERY

-0.19

ITES

-0.17

OTE

-0.16

er

-0.16

eur

-0.15

IENT

-0.15

 McKay

-0.15

inaire

-0.15

¶Į

-0.15

iel

-0.15

POSITIVE LOGITS

ension

0.57

ensi

0.54

ensively

0.54

ensions

0.52

ensible

0.52

ensive

0.52

ens

0.51

ended

0.50

ending

0.49

ense

0.43

Activations Density 0.020%

words related to comprehension and understanding, particularly in contexts where clarity or apprehension may be lacking

No Comments

No Known Activations