INDEX

Explanations

variations of the word "I" or "me" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

kok

-0.17

 EntityState

-0.16

ansa

-0.16

try

-0.15

pta

-0.15

oms

-0.15

kaz

-0.14

med

-0.14

anni

-0.14

ming

-0.13

POSITIVE LOGITS

imo

0.18

orda

0.17

esian

0.16

ises

0.15

esto

0.15

Ð¸Ð¼Ð¾

0.14

atural

0.14

_signals

0.14

place

0.14

 place

0.14

Activations Density 0.020%

variations of the word "I" or "me" in different contexts

No Comments

No Known Activations