INDEX

Explanations

references to the "Academy" and related concepts, potentially related to awards or recognition

oai_token-act-pair · gpt-3.5-turbo

references to the Academy Awards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

TPS

-0.78

ppings

-0.74

pd

-0.70

nir

-0.69

 Meier

-0.69

sheet

-0.68

 subsistence

-0.64

ï¸ı

-0.64

OPLE

-0.64

ulkan

-0.63

POSITIVE LOGITS

 Awards

0.98

emies

0.96

rats

0.81

 Ruins

0.80

æ©

0.79

 Award

0.78

Grade

0.77

 Academy

0.76

ctor

0.73

rylic

0.72

Activations Density 0.010%

references to the "Academy" and related concepts, potentially related to awards or recognition

references to the Academy Awards

No Comments

No Known Activations