INDEX

Explanations

instances where two entities or actions are connected or occurring together

oai_token-act-pair · gpt-3.5-turbo

phrases and terms associated with connections and relationships, particularly emphasizing the concept of conjunctions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lying

-0.72

gun

-0.66

cas

-0.65

ifts

-0.64

mbuds

-0.63

bern

-0.61

ambling

-0.61

cigarette

-0.60

 Citiz

-0.60

mac

-0.60

POSITIVE LOGITS

ually

0.90

ality

0.87

ioned

0.84

alities

0.77

ally

0.75

creen

0.74

SHIP

0.73

ivity

0.73

 conjunction

0.72

naire

0.70

Activations Density 0.008%

instances where two entities or actions are connected or occurring together

phrases and terms associated with connections and relationships, particularly emphasizing the concept of conjunctions

No Comments

No Known Activations