INDEX

Explanations

elements of a programming structure or code syntax

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Dann

-0.18

neum

-0.15

ãģŁãģł

-0.15

 Snap

-0.15

ÙĪÙĨÙĩ

-0.15

fund

-0.15

 Dive

-0.14

inch

-0.14

omain

-0.14

Ã´ng

-0.14

POSITIVE LOGITS

ALLERY

0.17

uder

0.16

 truth

0.14

Dice

0.14

icz

0.14

dice

0.14

akov

0.14

ahn

0.13

jedn

0.13

COPY

0.13

Activations Density 0.007%

elements of a programming structure or code syntax

No Comments

No Known Activations