INDEX

Explanations

underscores in textual data, indicating placeholders or special formatting

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.2.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

Ø±ÙĬØ±

-0.16

.ci

-0.14

woke

-0.14

strup

-0.14

midi

-0.13

ÑĨÐµÑĢ

-0.13

wner

-0.13

mong

-0.13

orny

-0.13

dbc

-0.13

POSITIVE LOGITS

atre

0.16

 minus

0.15

 Mutable

0.14

ureau

0.14

ÏĢÎ»

0.13

bard

0.13

amber

0.13

LETE

0.13

iev

0.13

-helper

0.13

Activations Density 0.033%

underscores in textual data, indicating placeholders or special formatting

No Comments

No Known Activations