INDEX

Explanations

instances of the prefix "dis-" used in various contexts

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

Not in Any Lists

No Comments

Negative Logits

tir

-0.19

stadt

-0.18

trad

-0.17

-0.16

cision

-0.16

td

-0.15

VF

-0.15

tempts

-0.15

tot

-0.14

 dismissed

-0.14

POSITIVE LOGITS

dis

0.21

son

0.19

ordered

0.19

ney

0.18

perse

0.18

ulf

0.18

continue

0.17

amb

0.17

 charge

0.17

colour

0.17

Activations Density 0.017%

instances of the prefix "dis-" used in various contexts

No Comments

No Known Activations