INDEX

Explanations

instances of the word "w" and its variations in Polish, as well as contextual uses of prepositions and conjunctions indicating place or time

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

 Asian

-0.16

hle

-0.15

avia

-0.15

alom

-0.15

udson

-0.15

eration

-0.14

weg

-0.14

ovich

-0.14

hong

-0.14

WL

-0.14

POSITIVE LOGITS

/cmd

0.17

yr

0.16

iat

0.15

aka

0.15

ad

0.15

raz

0.15

atur

0.14

ron

0.14

inka

0.14

ette

0.14

Activations Density 0.015%

instances of the word "w" and its variations in Polish, as well as contextual uses of prepositions and conjunctions indicating place or time

No Comments

No Known Activations