INDEX

Explanations

expressions of uncertainty or confusion about one's current state or circumstances

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

states of being or current conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

heny

-0.66

 Canaver

-0.65

umn

-0.62

idates

-0.61

ERY

-0.61

edIn

-0.60

 Inqu

-0.60

 Clarks

-0.60

Ov

-0.60

 Swim

-0.59

POSITIVE LOGITS

 supposed

1.22

 headed

1.00

 positioned

0.94

held

0.93

 portrayed

0.92

 able

0.92

 perceived

0.90

 located

0.90

 situated

0.88

 wont

0.84

Activations Density 0.164%

expressions of uncertainty or confusion about one's current state or circumstances

states of being or current conditions

No Comments

No Known Activations