INDEX

Explanations

frequent conjunctions in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.20.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

ilda

-0.18

-0.16

 block

-0.16

hin

-0.15

å¯

-0.15

olk

-0.15

uteur

-0.14

ERO

-0.14

ÏĦÎ·

-0.14

tin

-0.14

POSITIVE LOGITS

exus

0.16

lesi

0.14

 Policies

0.14

agged

0.13

ovie

0.13

anus

0.13

 SÃ¶z

0.13

 amat

0.13

itas

0.13

opa

0.13

Activations Density 0.002%

frequent conjunctions in the text

No Comments

No Known Activations