INDEX

Explanations

specific historical references and terms related to specific locations, events, or people

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

New Auto-Interp

Configuration

jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

torch.float32

Hook Point

blocks.12.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

HuggingFaceFW/fineweb

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 intersper

-1.60

 trouva

-1.13

 universale

-1.10

 exé

-1.08

 écl

-1.04

 soigne

-1.04

 espé

-1.04

 matel

-1.02

 tén

-1.00

 gouver

-1.00

POSITIVE LOGITS

Glej

0.92

arrol

0.81

polski

0.81

Și

0.79

sizePolicy

0.78

 Paglinawan

0.78

توضیحات

0.77

المصادر

0.76

Quiénes

0.76

Referencoj

0.75

Activations Density 0.793%

specific historical references and terms related to specific locations, events, or people

No Comments

No Known Activations