INDEX

Explanations

references to cooking or preparing food

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.28.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

 Salisbury

-0.16

åŃĿ

-0.14

tec

-0.14

 Salmon

-0.14

isko

-0.14

_rpc

-0.14

 Slav

-0.14

Ñī

-0.14

readcr

-0.14

dorf

-0.14

POSITIVE LOGITS

 sugar

0.77

 Sugar

0.71

Sugar

0.66

 sugars

0.59

sug

0.55

ç³ĸ

0.55

Sug

0.53

 glucose

0.52

 sweetness

0.51

 cane

0.50

Activations Density 0.088%

references to cooking or preparing food

No Comments

No Known Activations