INDEX

Explanations

terms related to medical treatments and interventions

oai_token-act-pair · gpt-3.5-turbo

references to health-related therapies and conditions

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

 whilst

-0.74

 Whilst

-0.73

Firstly

-0.65

 fantastic

-0.60

Advertisements

-0.58

 Zionist

-0.58

 unbelievably

-0.57

 fucking

-0.57

 brilliantly

-0.56

 glorious

-0.56

POSITIVE LOGITS

"'

0.95

 meanwhile

0.94

 countered

0.93

''

0.87

 also

0.87

ONSORED

0.82

"(

0.82

?,

0.79

 spokeswoman

0.79

Activations Density 0.905%

terms related to medical treatments and interventions

references to health-related therapies and conditions

No Comments

No Known Activations