INDEX

Explanations

phrases related to an ideal state or concept

oai_token-act-pair · gpt-3.5-turbo

expressions of the concept of "ideal."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

bane

-0.82

ktop

-0.77

rough

-0.73

been

-0.71

weak

-0.69

hani

-0.69

upon

-0.68

worthiness

-0.68

SEE

-0.67

SPONSORED

-0.67

POSITIVE LOGITS

istically

1.05

imates

1.03

istic

1.02

yip

0.86

 embodiment

0.85

imum

0.82

 ideal

0.80

igslist

0.78

 representation

0.77

ized

0.74

Activations Density 0.010%

phrases related to an ideal state or concept

expressions of the concept of "ideal."

No Comments

No Known Activations