INDEX

Explanations

prefixes and suffixes related to negation or reversal

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

YTE

-0.15

_marshall

-0.15

TINGS

-0.15

ãĥ¬ãĥ³

-0.14

ContentLoaded

-0.14

ãĥ³ãĥĶ

-0.14

ãĤ·ãĥ§ãĥ³

-0.14

âĹİ

-0.14

Iss

-0.14

ï¼ļ<

-0.14

POSITIVE LOGITS

athing

0.17

0.16

åĳ½

0.15

.setResult

0.15

meld

0.15

cre

0.14

zza

0.14

bage

0.14

 bild

0.14

 dilig

0.14

Activations Density 0.104%

prefixes and suffixes related to negation or reversal

No Comments

No Known Activations