INDEX

Explanations

names or titles associated with individuals or groups, particularly in a political or historical context

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

ember

-0.16

dap

-0.16

ahoma

-0.16

plex

-0.15

Ø·Ø§ÙĦ

-0.15

artz

-0.15

ÑĢÐ¾ÑģÑĤÐ¾

-0.15

 Ø§ÙĦØ±ÙħØ²ÙĬØ©

-0.15

eps

-0.15

.transparent

-0.14

POSITIVE LOGITS

oire

0.15

-Jul

0.15

 kidd

0.15

uden

0.14

xCF

0.14

ouden

0.14

-D

0.14

0.13

 Wenger

0.13

Ã¶yle

0.13

Activations Density 0.020%

names or titles associated with individuals or groups, particularly in a political or historical context

No Comments

No Known Activations