INDEX

Explanations

references to transportation hubs, particularly train and bus stations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ake

-0.15

ise

-0.15

ial

-0.14

ofile

-0.14

 planes

-0.14

ogn

-0.13

áº¡o

-0.13

andle

-0.13

ails

-0.13

 Steering

-0.13

POSITIVE LOGITS

 station

0.93

 Station

0.81

station

0.77

 stations

0.75

Station

0.68

_station

0.65

 Stations

0.64

.station

0.64

stations

0.58

 ÑģÑĤÐ°Ð½

0.56

Activations Density 0.112%

references to transportation hubs, particularly train and bus stations

No Known Activations