INDEX

Explanations

names of individuals involved in legal situations or criminal activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Diff

-0.16

 Diff

-0.16

orus

-0.16

MASK

-0.15

rou

-0.15

 automÃ¡t

-0.15

ren

-0.15

sle

-0.15

 SÃ¼

-0.14

ORTH

-0.14

POSITIVE LOGITS

alez

0.22

quez

0.22

ateg

0.22

endez

0.21

inez

0.20

ATEG

0.20

arez

0.19

Ð°ÑĤÐµÐ³

0.19

chez

0.19

 Lara

0.19

Activations Density 0.168%

names of individuals involved in legal situations or criminal activity

No Comments

No Known Activations