INDEX

Explanations

references to stairs and climbing-related activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.24.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

UIL

-0.15

ocket

-0.14

shima

-0.14

dyn

-0.14

 menus

-0.13

rÃ³

-0.13

 Dagger

-0.13

_matrices

-0.13

tls

-0.13

OKIE

-0.12

POSITIVE LOGITS

 stairs

0.66

 staircase

0.56

 stair

0.55

 steps

0.54

 Steps

0.47

 ladder

0.45

 climb

0.45

stairs

0.44

 escal

0.42

steps

0.42

Activations Density 0.119%

references to stairs and climbing-related activities

No Comments

No Known Activations