INDEX

Explanations

comparative and superlative adjectives indicative of performance or quality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

eniable

-0.18

sto

-0.16

æ¯Ķè¼ĥ

-0.16

plr

-0.15

æ¯Ķè¾ĥ

-0.15

efon

-0.15

pll

-0.15

atori

-0.14

æŁı

-0.14

ãĥ³ãĥģ

-0.14

POSITIVE LOGITS

 than

0.44

 Than

0.30

 then

0.30

 THAN

0.26

than

0.26

_than

0.25

Than

0.24

	then

0.20

 THEN

0.20

 Then

0.19

Activations Density 0.116%

comparative and superlative adjectives indicative of performance or quality

No Comments

No Known Activations