INDEX

Explanations

references to time, specifically occurrences of the word "last" and associated time indicators

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

iid

-0.18

les

-0.14

appa

-0.14

undi

-0.14

lets

-0.14

idente

-0.14

hid

-0.14

REFERRED

-0.13

let

-0.13

EMPLARY

-0.13

POSITIVE LOGITS

/current

0.19

-minute

0.19

most

0.17

cial

0.17

.fm

0.16

-last

0.16

veis

0.16

anza

0.16

itude

0.15

iferay

0.15

Activations Density 0.046%

references to time, specifically occurrences of the word "last" and associated time indicators

No Comments

No Known Activations