INDEX

Explanations

comparative words indicating an increase or comparison

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize an increasing quantity or degree

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

æ©

-0.76

OCK

-0.75

ï¿½ï¿½ï¿½ï¿½

-0.72

xtap

-0.72

borg

-0.71

IENCE

-0.67

 Accounting

-0.66

selves

-0.66

ï¿½ï¿½

-0.64

 Classification

-0.64

POSITIVE LOGITS

 than

0.90

 complicated

0.87

 expensive

0.86

 adventurous

0.84

 frequent

0.83

ls

0.81

 manageable

0.81

 affluent

0.80

 elaborate

0.80

 generous

0.80

Activations Density 0.035%

comparative words indicating an increase or comparison

phrases that emphasize an increasing quantity or degree

No Comments

No Known Activations