INDEX

Explanations

conditional phrases and comparisons related to performance and efficiency

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.7.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

 Newsp

-0.16

plemented

-0.16

-0.15

URA

-0.14

 INDIRECT

-0.14

bedtls

-0.14

egrity

-0.14

-0.13

POSITIVE LOGITS

any

0.18

/or

0.17

 anywhere

0.17

wig

0.16

amage

0.16

 anything

0.15

lock

0.15

agic

0.15

bom

0.14

//{{

0.14

Activations Density 0.098%

conditional phrases and comparisons related to performance and efficiency

No Comments

No Known Activations