INDEX

Explanations

images or actions described in the text

oai_token-act-pair · gpt-3.5-turbo

actions and states involving individuals 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

-0.89

 insure

-0.67

 guarantees

-0.65

fortunately

-0.64

 accordingly

-0.64

 criterion

-0.64

ItemTracker

-0.64

ktop

-0.63

 antiv

-0.62

flag

-0.61

POSITIVE LOGITS

ersive

0.75

 Grey

0.69

-+-+

0.69

 surrounded

0.69

 cruc

0.67

 weep

0.64

 engulfed

0.64

amorph

0.64

cled

0.63

unl

0.63

Activations Density 0.590%

images or actions described in the text

actions and states involving individuals in various contexts

No Comments

No Known Activations