INDEX

Explanations

references to categories or collections, particularly the term "all" 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 @ 9-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L9R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.9.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

ankan

-0.16

undan

-0.15

agem

-0.15

ontent

-0.15

stime

-0.14

æĹıèĩªæ²»

-0.14

ovit

-0.14

kker

-0.14

andum

-0.14

jvu

-0.14

POSITIVE LOGITS

 pace

0.17

igator

0.17

otre

0.16

usion

0.16

 Pace

0.15

StateManager

0.15

ties

0.14

iances

0.14

keley

0.14

igators

0.14

Activations Density 0.071%

references to categories or collections, particularly the term "all" in various contexts

No Comments

No Known Activations