INDEX

Explanations

phrases related to imagining or speculating about different scenarios or experiences

oai_token-act-pair · gpt-3.5-turbo

contemplative reflections about experiences and future possibilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Job

-0.59

=-=-

-0.58

 Transparency

-0.54

abad

-0.54

hement

-0.53

LIM

-0.52

arget

-0.52

rf

-0.52

CPR

-0.52

 Precision

-0.51

POSITIVE LOGITS

 like

0.90

like

0.74

 entail

0.71

 afterwards

0.69

 Like

0.68

 afterward

0.67

 tastes

0.67

ural

0.67

 tasted

0.66

 look

0.66

Activations Density 0.184%

phrases related to imagining or speculating about different scenarios or experiences

contemplative reflections about experiences and future possibilities

No Comments

No Known Activations