INDEX

Explanations

affirmative or confirming expressions and adverbs

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

inho

-0.16

ocre

-0.15

itol

-0.14

ailable

-0.14

ÃĹ↵↵

-0.14

rogen

-0.13

opsis

-0.13

astos

-0.13

 blown

-0.13

uchos

-0.13

POSITIVE LOGITS

has

0.17

ANNOT

0.16

-ÑĤÐ°ÐºÐ¸

0.15

 took

0.15

 went

0.15

 comes

0.15

uma

0.15

chÃ¡zÃŃ

0.15

 seems

0.14

vÄĽt

0.14

Activations Density 0.328%

affirmative or confirming expressions and adverbs

No Comments

No Known Activations