INDEX

Explanations

references to specific toy collections and their characteristics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 18-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L18R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.18.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

antz

-0.16

 nackte

-0.14

ksi

-0.14

pone

-0.14

punkt

-0.13

xbd

-0.13

stanov

-0.13

_PARTITION

-0.13

env

-0.13

hea

-0.13

POSITIVE LOGITS

 pose

0.16

Collect

0.16

 Collect

0.16

 collect

0.15

 figure

0.15

him

0.15

slaught

0.15

íıī

0.14

acic

0.14

longleftrightarrow

0.14

Activations Density 0.006%

references to specific toy collections and their characteristics

No Comments

No Known Activations