INDEX

Explanations

quantities or numbers

oai_token-act-pair · gpt-3.5-turbo

the phrase "how many" in various contexts

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

ital

-0.83

hern

-0.73

ocracy

-0.70

acus

-0.70

stead

-0.70

HF

-0.68

bringing

-0.67

olitics

-0.67

extremely

-0.66

00200000

-0.66

POSITIVE LOGITS

 times

0.93

 iterations

0.80

 servings

0.77

 parentheses

0.75

 calories

0.72

 millenn

0.71

 numbered

0.70

 repet

0.70

een

0.69

 dozen

0.69

Activations Density 0.013%

quantities or numbers

the phrase "how many" in various contexts

No Comments

No Known Activations