INDEX

Explanations

occurrences of the word "so" indicating emphasis or continuity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 4-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L4R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.4.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

Äĩ

-0.17

ÄįÃŃ

-0.15

tl

-0.15

missive

-0.14

getRow

-0.14

 Smoke

-0.14

suming

-0.14

 smoke

-0.14

blr

-0.14

-0.13

POSITIVE LOGITS

rane

0.16

gebn

0.16

eyer

0.15

pesan

0.15

 eslint

0.15

ÑĻ

0.15

Fav

0.14

opsy

0.14

Ð¾Ð»ÑĥÑĩ

0.13

Ø³Ø·

0.13

Activations Density 0.143%

occurrences of the word "so" indicating emphasis or continuity

No Comments

No Known Activations