INDEX

Explanations

words related to stating facts or information

oai_token-act-pair · gpt-3.5-turbo

statements expressing certainty or factual claims

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

TPPStreamerBot

-0.70

Eva

-0.69

Mos

-0.67

ummies

-0.65

Els

-0.64

Args

-0.64

abal

-0.63

ascript

-0.63

gently

-0.63

ushima

-0.62

POSITIVE LOGITS

 implication

0.63

 embold

0.63

 depends

0.63

 overriding

0.62

 aside

0.61

 varies

0.60

isn

0.60

quickShipAvailable

0.60

 downside

0.60

 arises

0.59

Activations Density 0.338%

words related to stating facts or information

statements expressing certainty or factual claims

No Comments

No Known Activations