INDEX

Explanations

occurrences of the word "prze" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.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

aret

-0.14

rack

-0.13

.sav

-0.13

vik

-0.13

iesz

-0.13

ady

-0.13

 foreigners

-0.13

amera

-0.13

ert

-0.13

POSITIVE LOGITS

 Gross

0.17

pad

0.17

 transitions

0.16

Transition

0.16

gon

0.16

transition

0.16

ãģĭãģĹ

0.15

_transition

0.15

 transition

0.15

du

0.15

Activations Density 0.008%

occurrences of the word "prze" in different contexts

No Comments

No Known Activations